A look back at ISC23 including quantum computing, EuroHPC, the future of Supercomputing with a backdrop of Integrated Research Infrastructure (IRI), AI, and Cloud, and whether we are living in times when “everybody” is a systems company.
This episode is sponsored by Lenovo.
The OrionX editorial team manages the content on this website.
We are trying something new, posting the transcript of a podcast. We start with one of the more popular OrionX Download podcasts on 5G. It was described as “probably one of the best discussions anywhere about 5G” and I obviously agree! 5G is a big deal as a technology and as a matter of geopolitical importance.
Welcome to The OrionX Download. This is a podcast where we discuss and simplify the big ideas in technology that are changing the world.
Intro: Hello everyone. Welcome again to The OrionX Download podcast. This is Shahin Khan and today I have the pleasure of speaking with Walt Maclay who is the president of Voler Systems. Voler Systems is a company that does electronic product design, development, risk assessment, verification, and they do it for connected smart devices. They’ve got a lot of skills in sensors and wireless technologies, and of course wireless technologies and 5G communication are big trends that we are tracking at OrionX. So without further ado, we’ll go to our conversation.
Shahin Khan: Walt, thanks for being here. Please tell us more about you and Voler Systems, and how radio communication in general and 5G in particular play a role in everything that you guys do.
Walt Maclay: Well, thank you for the opportunity to talk today. Voler Systems is a design house. We do electronic design and firmware, so we’re designing devices and we’ve got really good experience with sensors and wireless communication. So we do IoT and wearable devices. We’re always interested in the latest in wireless because wireless is changing all the time and 5G is currently the latest in wireless. It has a big impact on a lot of areas. As we’ll talk about as we go.
Shahin Khan: We know that 5G is the next generation of mobile cell phone, radio communication; it comes after 3G and 4G so we expect it to be better, faster, cheaper, but maybe before we drill down into all that it promises, maybe you can put that whole thing in perspective, just radio communication from the sensor all the way to the cloud and back.
Walt Maclay: How does all that work? There’s a whole variety of different types of wireless communication because there’s a whole bunch of different capabilities and needs, and basically what you’re trading off all the time is the speed that you’re sending data, how far you’re sending data, and how much power you use. Now if you’ve got something that’s plugged into the wall, the power isn’t such a big deal, but on a cell phone or a wearable device, you’ve got a battery and it’s a big deal. And then the speed is very important. Your cell phone needs to carry voice, and now it needs to carry video, and now with 5G we’re talking about way higher than that. And then there are the devices that may send the temperature once an hour. The data rate is dramatically different. So you’ve got different types of communication. Bluetooth LE is extremely low power and can send the data 10 or 20 feet reliably to a cell phone. It sends fairly slow data and for many applications, which is perfect. If you want to get directly to the Internet and you want to have high speed data, 5G is the way to go. You’ve got a lot of different choices and that’s why you do
Shahin Khan: Now when you go through all of these different protocols, ZigBee, BLE you mentioned, traditional WiFi, and then when you go farther distances 4G or now 5G. I imagine that in a typical deployment, a whole number of these are at play, that even if you go with 5G, maybe the backend is going to be wired or fiber or whatnot.
Walt Maclay: Oh yeah. When you go to the internet, once you get to the backbone of the internet, it can go through optical, it can go through wireless, it can go through telephone lines. There’s all kinds of different things involved. It’s a very complicated infrastructure. But for the consumers, really it’s about getting to the cell tower. That’s true pretty much with everything. When I talk about Bluetooth LE, or even WiFi, it doesn’t go directly to the internet. It doesn’t go directly to a cell tower. You have a hotspot. Well, the cell tower is like a hotspot, except it’s provided by someone else that’s public and they’re everywhere. You can roam from one to the other. They’re almost everywhere. So the ideal situation for many applications is to transmit directly to a cell tower or a hotspot of some kind and to be able to roam and have them everywhere. Well, 5G, and 4G as well, do exactly that. But there are some others. There’s NB IOT and LTE M which are cellular based, but they’re lower power and they’re designed for lower data rates. These are complimentary to 5G and they go similar distances. They go many kilometers or many miles.
Shahin Khan: So let’s start with frequency spectrum because it’s hard to talk about 5G without getting to that pretty immediately. And it seems like a lot of what 5G promises are enabled by just the frequency bands that are available.
Walt Maclay: Exactly. Yes, we have to talk about some details to make it all clear. So there are three different bands. Two of them are already in use in 4G. There’s the band that’s below one Gigahertz and the band that is between one Gigahertz and six Gigahertz below. One gigahertz has a long range. They claim up to 20 miles. But I don’t know about that. That’s in ideal conditions. So if you’re out in the country and you put up a cell tower and aim it down a freeway, it’ll go for a long distance. And then distance the one to six Gigahertz is pretty good. You’re used to seeing it on your cell phone. You often don’t know whether you’re in the low band or the mid band and it works pretty well, not quite as far distance. What’s new in 5G is the millimeter wave band. This is above six Gigahertz, and it’s opening the band that goes all the way to 89 Gigahertz. Currently it’s primarily in the U.S. In the 20 to 30 Gigahertz range. This band has some big advantages. It’s new space that’s not being used. The lower bands below six gigahertz are getting pretty used up. Also, you can have much higher bandwidth, so it provides much higher data rates, so when they talk about high data rates, they’re talking primarily about the millimeter wave band. The downside of the millimeter wave band is it doesn’t travel very far and we’ll talk more about that
Shahin Khan: And in fact, speaking of that, really none of this matters. If it’s not available, what does the coverage look like now and how do we expect it to change over time? Who’s offering what
Walt Maclay: Coverage is very dynamic right now, I’ve got data from November of 2019 which is now quite obsolete. Then there’s some more recent data, but as of November, the major cities were covered with 5G and if you take California as an example, three of the four carriers didn’t cover the Bay Area. I think that they do now, and I don’t try to keep track on a month by month basis. It’s being built out very quickly. In the country, however, I think you’re going to find the availability is much less now. I believe that if you get a phone, pretty much any phone that’s 5G it will work on 4G. So if you buy a 5G phone, you’ll still have service. You just may not get any benefit from the 5G capability. So if you pay extra, what have you gotten?
Shahin Khan: Okay, so now 5G in addition to cell phones there’s obviously an IOT connection and it seems like there are classes of services that they categorize. What are those?
Walt Maclay: This is what’s really new about 5G. There are three different types of service that are very different.
The first one is enhanced mobile broadband or eMBB. This is the standard cell phone service that goes to computers, laptops, iPods, and whatever. This is going to give you much higher data rates up to two gigabits per second. Much faster than what you are used to.
The second one is ultra reliable, low latency communication or URLLC. This is a completely new category. Typically these will be on premises devices. For example, they’re used for a factory. Within the factory you can set up your own self service and cover your factory and have extremely fast, very low latency and very reliable communication wireless, but it could also be in a stadium for everybody to use their cell phones.
The third one is completely different. Massive machine type communications mMTC. This is really IoT; very low data rates. Latency is unimportant, but power is everything and it’s completely different from the others, and the thing is that they haven’t even finished the standards for massive machine type communications or IoT. What’s available today is NB-IoT and LTE-M which have been built out in 4G cell towers. Today, IoT hasn’t really changed on 5G
Shahin Khan: When we talk about inside the premises, as you mentioned in a factory, why not just use WiFi? Why do you need to go 5G inside the factory?
Walt Maclay: Good question. WiFi is also improving and both WiFI and 5G are using new technology that’s providing more devices in an area and lower latency. I’m not aware of the ultra reliable low latency communication being offered from WiFi. So it seems that there’s something special there. But yeah, WiFi is definitely a competitor for this type of application.
Shahin Khan: Right, because sometimes I think just the combination of BLE and WiFi should be good enough for a whole lot of use cases, and then maybe LTE-M to get out there and if all that works, then why make such a big deal about 5G? But we’re going to discuss that. Okay, so bandwidth, everybody thinks of 5G immediately as it’s just going to be faster data communication. We know it depends on the frequency, but what can people expect to get and how is that different from geography to geography?
Walt Maclay: So the bandwidth you get depends entirely on which service you’re getting and what frequency band you’re in. If you’re in the sub one gigahertz band, you know they’re re-purposing 4G to 5G for that band, you’re going to get a modest improvement over 4G. You may not even notice, You’ll get something like a hundred megabits per second going one direction, 50 megabits per second going the other. In the one to six gigahertz band, it’s going to be significantly different. You can get up to 400 megabits per second, which is perhaps four times what you can get or three times what you can currently get in 4G – significantly different. But the really big improvement is in the millimeter wave band where you can get two gigabits per second. This is 20 times what you can get in 4G; a huge difference.
Walt Maclay: There’s a big issue there though. The millimeter waves don’t pass through buildings. They don’t pass through windows. If you put your hand in front of it, you block it. It doesn’t go through rain or snow. And the distance that it travels is a small fraction of a mile, so you can only find these in dense urban areas. You’re never going to have the millimeter wave band, as far as I can tell, in a rural area, and I don’t see how it can even be in suburban areas, because you’ll need a cell tower approximately every hundred to 300 meters to get the service and we know that there’s been a problem in putting lots of cell towers into suburban areas.
Shahin Khan: It’s already hard to have the ones that you already have there. Exactly,
Walt Maclay: And making them 10 times more dense or a hundred times more dense; I don’t see it happening.
Shahin Khan: Right, and maybe we’ll touch on it later, but of course there’s a whole lot of controversy about cell towers in general. Yes, let’s move to latency now. Latency is another promise of 5G. How does that look?
Walt Maclay: 5G offers very low latency, but we’ve got to talk about the conditions. For some applications latency is fine already. For voice it is fine. Even video. If you’re just downloading a TV show 4G works and 5G will work fine, but where latency is really cool: Let’s say you’re doing an AR or VR application, and when you turn your head, it needs to download the video instantly to show you video where you’re looking. Delay of tens of milliseconds is noticeable. 5G can get down into the one millisecond latency.
Walt Maclay: However, think about how far away the other person is. Let’s say you’re playing a game. If they’re on the other side of the country, it takes, well 186,000 miles per second. It’s about 30 milliseconds round trip across the country. You know that you’re not going to get one millisecond. You have to consider round trip time. When they talk about the latency, they’re talking about one direction going just from the transmitter to the receiver. So say it’s from your phone to the cell tower. It doesn’t count the time coming back. It doesn’t count the time going through the internet itself. And every time there’s a switch on the internet, you add about a millisecond. So in many applications you’re going to be 20, 30, or 40 milliseconds. In 4G you got maybe 50 to 100 milliseconds, so it’s better, but maybe not dramatic.
Walt Maclay: If you’re in those ultra reliable low latency applications were it’s just on a factory floor and is traveling a hundred meters, then you can get to one millisecond latency because the distance is very small and you’re not going through the backbone of the internet. So in some applications you can get it. I see one application would be for remote surgery. Across the country you’ve got a bit of a problem, but if it’s not too far. If it’s just in a rural area outside the city, you can get a very fast and very reliable connection, if you have the high frequency millimeter wave cell tower available in this rural location. The hospital would have to set up a cell site just for the application. So this would be a case where the local hospital has the tower and some remote surgeon is actually doing the surgery. There’s a shortage of surgeons in the country, but if you can get somebody in the city to do it, that would be a nice.
Shahin Khan: So two things come across to me there. One is interactivity and one is access to some remote expertise or talent. That is compelling. Yeah.
Walt Maclay: So having that remote is good, but with surgery having anything more than a 10 millisecond or so delay, you begin to notice that you move something and it didn’t seem to move when you moved it and it really interferes with your ability to move. You know, if you move your hand and your hand doesn’t move when you said to move it, your brain doesn’t operate very well with that.
Shahin Khan: It feels like those movies when the dubbing is not synced up with the audio, so the perception is that poor,
Walt Maclay: Your perception is on the order of tens of milliseconds. If it’s 10 milliseconds, you’re fine. At 50 milliseconds it’s very noticeable.
Shahin Khan: Interesting. Interesting. Very good. So we kind of had the similar jump going from 3G to 4G. I remember that and that really did not seem to be a big deal. Walt, would you paint a picture for us of what’s going on, on the standards side and summary of what all is being promised.
Walt Maclay: Okay, that’s a big topic. So first, if all you did was use your cell phone to do voice, when you went from 3G to 4G you didn’t notice any change. But there are services that are provided that rely on 4G that have made things very different. You have the Internet on your phone in real time, where prior to 4G you didn’t. It has made a big difference, but it happened maybe just a little slowly and people didn’t really notice. But some people noticed, and it turned into billion dollar industries. There are startups where their business is possible because of 4G.
5G is probably going to do the same thing, but what they’re promising isn’t always available. You’re not going to have the millimeter waves in the country or probably not in suburbia. So those high speeds, and low latencies probably aren’t going to be available to you there.
Walt Maclay: Even in the dense city, if you’re indoors, the only way you can receive millimeter wave signals is to have a pico-cell. You may be familiar with pico-cells in 4G, they are handheld devices. If you’re in a busy area like a trade show, you want to have your own cell tower to guarantee that you can have cellular, so you bring in a pico-cell.
Well, you’ll be able to do the same thing with 5G. The antenna has to be outside because the signals don’t go into the building, but with that, and if you put one in every room, you can have 5G indoors in the millimeter wave band. Right now these are pretty expensive. They may be in the thousand dollar range, but they’ll come down. There are some things that have to be done and I suspect that a lot of this will happen and we’ll start to see 5G in those places in dense urban areas – the millimeter wave 5G that gives you up to two gigabits per second. They claim higher, but I haven’t seen higher yet.
Shahin Khan: You’re saying that it’s going to be hard to get everything that 5G offers at the same time?
Walt Maclay: Yes. For example, I forgot which magazine did this, but they went out and did some studies. They got some measurements of speed. It was with AT&T. They had AT&T take them to a place where they could get 1.8 gigabits per second. But they pointed out that when they were more than 100 or 300 feet away, so about 100 meters, they didn’t really get the same performance. So that means you’ve got to have a cell tower within about a hundred meters or less to get millimeter wave signals, and you can’t be indoors. You can’t be in your car. It can’t be snowing or raining. So this is very limited for right now.
Shahin Khan: Let’s talk about use cases. You already mentioned augmented reality, virtual reality. What else seems like a reasonable legit use case?
Walt Maclay: Well, for the ultra reliable low latency, there’s some fascinating use cases. I mentioned a few of them like remote surgery, use in a factory, and another one is on power distribution. Another use case is to send the signal down the power line for many miles. It’s just along the power line, and again this is on premise, this is not available to cell phone users. This is available just to the utility, but it will allow them to monitor and control the power lines better than they do now. This is a big deal. It’s all transparent to most people, but it’s a big deal.
Another one that’s more visible is in automobiles where 5G will go 2 kilometers down the road. It will be picked up by the car. The car will be able to communicate with the road and with other cars. You’ll be able to see around the corner traffic lights will be controlled to improve the flow of traffic. This is a smart city application. This could be a really big deal, particularly with self driving cars. And again, that’s not going to be picked up by your cell phone. That’s going to be picked up by your car.
So there are some very interesting things that are coming up. Another one would be if you get a large group of people together in a stadium or an entertainment facility, you can have the millimeter wave band in that facility and everybody can have communications at very high speeds and very low latencies through their cell phones and other devices. Who knows what kind of new interesting things will happen. Probably in the realm of entertainment. Entertainment is a really big area and so I expect a lot to happen in that area.
Shahin Khan: Well, you’re right, if you’re sitting in a stadium and you can just watch a replay on your own phone from five different angles, that’s new.
Walt Maclay: That could happen. But what will be most interesting will be the things we can’t even think of.
Shahin Khan: All right. It’s now coming back to the end-to-end part of it. What do you need in an end-to-end way to have a real 5G infrastructure in place? And where I’m going with this is maybe not to the whole geopolitical aspect of this. Who makes these devices starting from the sensor all the way up. What is really interesting to me is the kind of vendors that are in this market that some of our audience may not have heard of besides the usual well known players.
Walt Maclay: Okay. There’s a huge infrastructure that’s providing 5G. You start out with the network suppliers, everybody’s heard of AT&T, Verizon, T-Mobile, and Sprint. There are the phone manufacturers, Motorola, Apple, LG, and many others. And then there’s equipment manufacturers. They make equipment that’s used, for example, in the cell towers, Ericsson, Nokia, and many others. Then there are the companies making the semiconductors to go into the equipment, Qualcomm, Intel, Marvel, ARM, AMD, Xilinx, and many others. There are others beyond that. You could get into all kinds of things like sensors. For example, services. There are companies, that only make IP. They license technology that goes into all of these things. So there are hundreds or thousands of companies that are involved in 5G.
Shahin Khan: You know, I’m just gonna make a comment and that may very well be why your business is so critical. If you were just the average manufacturer and you’re trying to get into the IoT business, just making sense of all of that requires folks like your team who really know what they’re talking about in this area.
Walt Maclay: Yes. If you just want to use a phone, you go to AT&T, Verizon, T-Mobile, or Sprint, and they’ve worked with all these other companies to provide you the service. So it’s very simple. We do the same sort of thing. You want a device that connects in a certain way. We can give you the advice on what technology works best for your application.
Shahin Khan: A few other things we haven’t talked about. One of them really is device density and you keep reading that the spec for 5G is 1 million devices per square kilometer. Is that real? And wasn’t that kind of solved already with just a whole bunch of Wi-Fi hot-spots in one place?
Walt Maclay: Wi-Fi isn’t going to get you that kind of density. That’d be an awful lot of hot-spots interfering. 5G does claim a million per square kilometer now because some of the claims aren’t entirely real, I take it with a grain of salt, but they have done some things that do make this possible. For example, the antennas are directional, so you’ve got all the bandwidth in one very narrow direction and then just a few degrees over, you’ve got all the bandwidth again in a different direction. Well that dramatically increases the number of devices that you can have.
They also have not only split up the bandwidth by time, they split it up by frequency. So they’ve done a lot of things to increase the amount of devices that you can have now, even if it’s a lot less than a million per square kilometer. For right now. What that means is when you are traveling on 5G, you should not see the drops that you see with 4G.
Shahin Khan: I see. Because there’s just enough capacity and there are so many different ways to get to the tower. Right? Yes. I see.
Walt Maclay: Now if you move out of an area that has 5G, you might see a drop.
Shahin Khan: Right. But back to the whole stadium thing, how are they solving it now when without 5g you can get 4g in a stadium.
Walt Maclay: They have to have a lot of cell towers or cellular sites to handle the service. You’ll be able to do it with a fewer number of sites with 5g, but what you can’t do now is you can’t get to two gigabits per second speed and the low latency and in something like a stadium, that’s what’s going to make a big difference.
Shahin Khan: Right, right. Okay. Along those lines is where is the device, the whole global positioning and location tracking and the precision with which you can do that and 5G promises a lot in that area too.
Walt Maclay: That’s right. So location of a device is a big deal. There are all kinds of capabilities and services that require location and then it’s a big deal that you may not want people to know where you are. You can turn it off. So with 5G they claim they’ll be able to locate you within about a meter. Now, whether it’s an absolute meter or just a meter this way or that relative, a meter is pretty precise.
But if you think about it, you already know which direction someone is because you’ve got this directional antenna. On top of that they’re measuring signal strength, which gives you a pretty good indication. Though signal strength can change rapidly without a change in location. And then there’s a certain amount of triangulation. And now we’re going beyond what I know, but instead of just knowing that you’re within the cell tower range, which is within a mile or two, they’re going to know very close where you are perhaps better than with GPS.
Shahin Khan: Interesting. Interesting. And I think it’s also, I remember in the early days of cell phones when they didn’t have GPS, they would have to really rely on whatever towers were around you and try to figure out where you might be with some level of precision, but far from what GPS gives you. So this thing is going to get close to GPS or better, perhaps better. That’s impressive.
Walt Maclay: Now, depending upon which band you’re using, it may work indoors, and it may be able to tell you which floor you’re on. If you’ve got an emergency, and all we know is that you’ve got a phone in a certain location in the city, they don’t know which building you’re in or which floor you’re on. You could be dead before they find you. Now they’re going to know exactly where you are. So this could be a big,
Shahin Khan: Oh, so this is the stuff you used to see in movies and they now have it for real. Yeah. Interesting. Let’s talk about power. You mentioned that a big part of this is how far you go and how much data and how much power.
Walt Maclay: Yeah. So one of the things that 5G advertises is low power. Well, you can’t get everything for free. When you’re going really high speed and long distances it’s not low power. The IOT service will be low power, but that depends entirely on how much data you’re sending and how often you’re using it. So the 5G itself is not really low power. In fact, when you’re in the millimeter wave band, you’re using more power. One issue they’ve got with the 5G cell towers is they use about three times as much power as 4G, so they’re not so green, so there’s some things people complain about.
Shahin Khan: Right. I want to mention that we were both in a meetup recently that talked about power harvesting and I just wanted to mention the concept to our listeners just so they know that is going on as well, if you might speak to that a little bit.
Walt Maclay: Yes. Energy harvesting is fascinating and it’s growing. It uses sources like temperature differences, solar power, motion, and RF energy, for example. It pulls energy out of the air or out of the environment and converts it to electrical energy. It’s a great idea to reduce the need for batteries. The biggest limitation in energy harvesting is that it generates generally microwatts of power. Using solar power you can get more, but you might need a pretty big solar cell, so you’re not going to see energy harvesting replacing the battery or even recharging the battery on your cell phone unless you’re barely ever using it.
Shahin Khan: But it’s definitely an interesting addition and it’s free.
Walt Maclay: It’s a real interesting space. If a device doesn’t send very much data; if the sensor doesn’t use much power; energy harvesting is very interesting.
Shahin Khan: Right. I always imagine a vending machine in the middle of the desert or something.
Walt Maclay: Yes, a device in the desert can have a pretty good sized solar cell. So yes, it can easily run on energy harvesting.
Shahin Khan: Okay. The most important thing we haven’t talked about is cost. And I know that as an engineering shop, that’s exactly what drives all the trade offs that one must do in a design. So let’s talk about that and how that limits or opens up new markets.
Walt Maclay: Well, so if you look at 5G, I haven’t priced out phones, but I’m sure you pay a premium to get 5G. There’s more technology. The actual cost to the manufacturer is a bit higher, and I’m sure right now they’re making a higher margin so it’s significantly more. But take a look at something like IoT. I mentioned that in fact, IoT on 5G, they’re not really providing anything that wasn’t already here, but the cost for doing IoT direct to the internet has been dropping. A couple of years ago it was several dollars per month to have that service. It’s gotten down to well below a dollar a month depending upon the amount of data, but if you’re sending thousands of bytes per month, that’s trivial, so it’s very low cost.
As the price goes down, new uses come up, for example, wearable devices can now communicate directly to the Internet where one or two years ago the coverage wasn’t there. The only way to do it was you used Bluetooth or WiFi going to a phone and the phone would go to the Internet. Having a wearable device that requires a phone is often inconvenient. Now it can communicate directly.
Shahin Khan: Excellent. Very, very fascinating. Well, 5G continues to be a big mega trend of our times and really important in many ways. Before we close, maybe we can touch a little bit on the health care concerns of 5G. We mentioned that earlier in the program, that there is a lot of controversy, there’s a lot of misinformation. What is your perspective on all of that?
Walt Maclay: Well, there are people who are concerned about the millimeter wave frequencies. They haven’t been well studied for their effects on humans. If you remember a number of years back, the same issue was discussed about the lower frequency bands in any cellular phones, and in fact there are requirements to minimize the amount of radiation absorbed by the head when you’re using a cell phone.
There have been many, many studies, some of them paid for by people with an interest. So some of the studies are suspect. The problem is probably not a big deal, but there’s still uncertainty even in the lower bands. And I think in the higher bands, the same thing’s going to happen. It’s a trillion dollar industry where people want the results to come out a certain way. So there are going to be many studies that are slightly biased and you may not even know that they’re slightly biased, so we’ll never get this controversy to go away completely. There’s concern and I think there will continue to be concern for some time.
Shahin Khan: Yes. In agreement with you. I just don’t think there’s enough data and the data that is there seems just insufficient. That’s hopefully something that the industry will address.
Walt Maclay: Well the industry may try to address it, but if the industry backs a study that study is suspect.
Shahin Khan: That’s right. Well, I don’t know if I’ll include this, but I’ll share this quote from one of the VCs that said, no conflict, no interest! Very good. Thank you, Walt. Thank you for making the time and this very informative session. I look forward to staying in touch and track this thing as it progresses.
Walt Maclay: I have enjoyed the opportunity to talk and it’s a topic I enjoy talking about.
Shahin Khan: Perfect. Thank you.
Outro: That was great.
I want to thank Walt Maclay again for being with us and taking us through this very, very interesting conversation.
Before we end, I want to summarize what all 5G promises: it promises speed, low latency, high density in terms of number of devices per square kilometer, better precision in terms of positioning and location tracking, low power, high coverage, and a whole lot of backend efficiencies, including the three classes of service that we talked about.
On the other hand, you’re not going to get all of the above at the same time, or if you do, it’s going to be a very special situation. And then there are still lingering questions about its health effects, which in my opinion is entirely because there’s just not enough data, there’s not enough testing and not enough data to convince everybody, which is unfortunate because there’s also a lot of misinformation that can get in the way of the adoption of the technology unjustifiably.
So, with that, I look forward to the next one. Thank you for being here.
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.
Today, June 14, 2019, we released the second biannual list of Top 50 cryptocurrency mining pools.
We do this in conjunction with the Top 500 supercomputing list that is released twice a year, in June and November. That list has been a matter of national pride for the US, Japan, China, and many other countries.
Cryptocurrency mining is a specialized form of supercomputing, producing billions of dollars of economic value per year.
In the Information Age, money has become information. Bitcoin is energy converted to information and encapsulated as secure immutable transactions on a time chain. This is money in the Internet, that we call Money 3.0. Currently it is primarily a store of value, a sort of digital gold, but it continues to grow use cases as a medium of exchange, and unit of account.
Cryptocurrency mining operations are large-scale, run on clusters, but also consist of highly decentralized pools that anyone can join and contribute their equipment to the effort, for proportionate rewards. Most mining is done on custom ASIC computing rigs, highly optimized for the relevant crypto consensus algorithm.
Using statistics readily available on the hashing rates and block production rates for the large mining pools, we can tabulate the economic value produced by these pools.
We consider only mined coins, that is, those that use some type of Proof of Work algorithm such as Bitcoin’s Nakamoto consensus.
We do not consider coins created with other types of consensus mechanisms, since they require no significant supercomputer-class computation. This includes coins produced through premining, Proof of Stake, distributed Byzantine Fault Tolerance and the like since supercomputing resources are not involved.
While there are a number of lists that provide hash rates and block production rates for pools mining a single coin, our lists are the first aggregation of which we are aware.
This raises the question as to how to compare mined coins that have radically different hashing rates and whose consensus algorithms, although often similar to Bitcoin conceptually, differ in the details.
We settled on the economic value of the mined coins that are produced. This enables us to make comparisons across coins when rank ordering the list of mining pools.
We compare the dollar value of a day’s mining from a given pool, with that of other pools, across the top eight mined cryptocurrencies.
The top 10 mined coins have market caps above $0.5 billion dollars, and the #1 coin, Bitcoin, as of our snapshot taken on May 30, 2019, had a market cap of $154 billion.
When we rank order the top 50 mining pools we find that the top eight mined coins in economic value are: Bitcoin (BTC), Ethereum (ETH), Litecoin (LTC), Bitcoin Cash (BCH), Zcash (ZEC), Bitcoin SV (BSV), Dash (DASH), and Monero (XMR). All of these except Monero are ASIC-friendly, and production is dominated by ASIC miners and clusters. Monero relies on GPUs.
For Bitcoin, Ethereum, and Litecoin we have used 30 day averages as of May 30, 2019 for block production and hash rates; for the other coins 7 day average data was available.
Table 1: Top 8 Mined Coins (all mining pools, not just Top 50)
| Coin |
Hardware class |
Algorithm |
New coins / day |
Hash Rate |
Hash units |
Price 5/30/2019 US$ |
Economic Production per Day, Million $ |
Extrapolated Annual Production Million $ |
| Bitcoin |
ASIC |
SHA256 |
1800 |
47.1 |
Exa |
8701 |
15.662 |
5,717 |
| Ethereum |
ASIC |
Ethash |
13,600 |
172 |
Tera |
284 |
3.862 |
1,410 |
| Litecoin |
ASIC |
Scrypt |
14,825 |
352 |
Tera |
117.6 |
1.743 |
636 |
| Bitcoin Cash |
ASIC |
SHA256 |
1800 |
1.36 |
Exa |
469 |
0.844 |
308 |
| Zcash |
ASIC |
Equihash |
7200 |
4 |
Giga |
86.9 |
0.626 |
228 |
| Bitcoin SV |
ASIC |
SHA256 |
1800 |
2.03 |
Peta |
222 |
0.400 |
146 |
| Dash |
ASIC |
X11 |
1693 |
1.68 |
Peta |
172.2 |
0.292 |
107 |
| Monero |
GPU |
CryptoNight |
1934 |
329 |
Mega |
95.1 |
0.184 |
67 |
| Totals |
23.61 |
8,619 |
From Table 1 above, which is across all pools, not just the Top 50, we see that total annual economic value run rate (extrapolated from the recent average daily values) is about $8.6 billion. About 2/3 of the economic value created is from Bitcoin production alone, with about $15 million produced per day recently. Ethereum amounts to around one-quarter of that at almost $4 million per day. The next six coins add another $4 million daily. Overall around $24 million per day is currently being mined from all pools.
The locations of top mining pools can be multi-country. The next Table summarizes the major host countries for the Top 50 pools; China, the US, and Hong Kong account for 70% of the top 50 pools and almost all of the top 10 operators. China alone is responsible for nearly half of the annual value produced by the Top 50 pools. The Mixed category includes various combinations of US, China, the EU, Russia, or other Asian or European countries. This category has grown as Chinese operators begin to move to other geographies, as a result of pressure from the government to constrain cryptocurrency mining in China.
Table 2: Host Countries, Top 50 Pools
| Country | # Top Pools | Daily M$ | Annual M$ |
|
China |
18 |
10.717 |
3911.7 |
| US | 11 |
4.77 |
1742.5 |
| Hong Kong | 6 |
2.77 |
1009.6 |
| Mixed | 12 |
2.69 |
980.4 |
| Other | 3 |
1.18 |
430.0 |
| Totals | 50 |
22.12 |
8,074 |
Table 3: Top 10 Pool Operators (aggregated across coins)
| MultiPools | Coins | Number | Daily M$ | Annualized M$ | Country |
| BTC(dot)com | BTC, BCH |
2 |
3.06 |
1115 |
China |
| F2Pool | BTC, ETH, ZEC, BSV, LTC | 5 |
2.76 |
1007 |
China |
| Antpool | BTC, LTC, ZEC, BCH, DASH | 5 |
2.38 |
868 |
Hong Kong |
| Poolin | BTC, ZEC, LTC, BSV | 4 |
2.26 |
825 |
China |
| SlushPool | BTC, ZEC | 2 |
1.62 |
592 |
US |
| BTC.Top | BTC, LTC, BCH | 3 |
1.47 |
537 |
China |
| ViaBTC | BTC, LTC,BCH | 3 |
1.34 |
488 |
US |
| Huobi.Pool | BTC, ETH | 2 |
0.69 |
251 |
China |
| NanoPool | ETH, XMR | 2 |
0.50 |
182 |
US, EU, Asia |
| Bitcoin(dot)com | BTC, BCH | 2 |
0.34 |
124 |
US |
| Totals | 30 |
16.41 |
5,990 |
We have aggregated, for the top 10 operators, their results across all of the top eight coins, and summarized in Table 3 above. Some operators mine two different coins, others mine as many as five of the top eight. These pools account for, when broken out by coin, 30 of the entries in our Top 50 list.
The #1 operator is BTC.com based in China, and it produces $3 million a day of economic value. F2Pool, Antpool, and Poolin each produce over $2 million of cryptocurrency per day. These large operators are responsible for $6 billion of the $8 billion annual production by the top 50 pools. Three of the five largest operators are in China, one is in Hong Kong, and one is in the US.
The winners in this race, for this second list, are Bitcoin, naturally, with BTC.com again as the top pool, and China as the host country for the most top mining pools, including both #1 and #2 positions. Hong Kong has the #3 pool. The US has the second largest number of mining pools.
The economic value of mining has increased substantially. In the first list of November, 2018 we looked at the Top 30 pools, responsible for some $5.5 billion of annual run rate of mining. This new list of Top 50 pools indicates $8.1 billion of annual cryptocurrency creation (even the Top 30 for this list amounts to well over $7 billion).
We intend to update this list again in November, 2019. Suggestions and comments may be sent to: stephen.perrenod@orionx.net
A presentation with the full Top 50 list is available at SlideShare.net:
https://www.slideshare.net/perrenod/top-cryptosupers-201906v1-149547256
References:
Overall: coinmarketcap.com, coinwarz.com, cryptoslate.com
BTC: btc.com
ETH: btc.com, etherscan.io
BCH: btc.com, cash.coin.dance
LTC, ZEC, XMR, DASH: miningpoolstats.stream
Cryptocurrency topics: orionx.net/blog
Stephen Perrenod has lived and worked in Asia, the US, and Europe and possesses business experience across all major geographies in the Asia-Pacific region. He specializes in corporate strategy for market expansion, and cryptocurrency/blockchain on a deep foundation of high performance computing (HPC), cloud computing and big data. He is a prolific blogger and author of a book on cosmology.
Thanks to the efforts of my colleague Dr. Stephen Perrenod, OrionX launched a new CryptoSuper500 list last November, coinciding with the eagerly awaited TOP500 list of most powerful supercomputers. The TOP500 has spawned variations that look at different workloads and attributes, for example, the Green500, Graph500, and IO500 lists. CryptoSuper500 was inspired by those lists.
Please see the material for the inaugural edition of the CryptoSuper500 list here and here. As you can see from what we can share publicly, we track business, technology, and policy issues surrounding emerging technologies and take a realistic approach to what they can do, and under what circumstances their use is a wise choice. Blockchain and cryptocurrencies are very much in that mix as are IoT, Quantum Computing, 5G, and Cloud.
Cryptocurrency mining operations are often pooled and are very much supercomputing class, typically using accelerator technologies such as custom ASICs, FPGAs, or GPUs. Bitcoin is the most notable of such currencies. Scroll down for the top-10 list and see the slides for the full list and the methodology. Contact us if you’d like discuss it in person.
We issued the following press release to formally announce this 2nd edition. Read it below or here.
MENLO PARK, Calif. – June 14, 2019 – OrionX Research today released the second edition of its CryptoSuper500 list. The list recognizes cryptocurrency mining as a new form of supercomputing and tracks the top mining pools. Cryptocurrency technologies include blockchains, consensus algorithms, digital wallets, and utility and security tokens. Aided by a recent surge in cryptocurrency prices, the top 10 mined coins have market caps above $0.5 billion dollars, and the #1 coin, Bitcoin, as of our snapshot taken on May 30, 2019, had a market cap of $154 billion.
Cryptocurrency Supercomputing refers to large-scale cryptocurrency mining operations which are typically powered by accelerator technologies such as GPUs, FPGAs, or custom ASICs. Bitcoin is the most notable of such currencies. Cryptocurrency mining via Proof of Work is important since it continues to represent the most effective consensus algorithm to maximize security in a decentralized manner.
“Driven by Artificial Intelligence (AI), Blockchain, Internet of Things (IoT), and 5G connectivity, a new wave of enterprise applications demands supercomputing power. Cryptocurrency mining is an important part of that mix, impacting investment decisions globally,” said Dr. Stephen Perrenod, OrionX Partner and Analyst and the developer of the CryptoSuper500. “CryptoSuper500 tracks supercomputer class systems that are used for such mining applications.”
The annual economic value (AEV) of the top 5 pools in the current list compared to the last list (published in November 2018) are as follows:
ViaBTC in the US and BTC.Top in China were among the top 5 in the previous list and have been edged out by F2Pool and Poolin in the current list.
The eight coins mined with Proof of Work consensus algorithms and that made the cut for our analysis are Bitcoin, Ethereum, Litecoin, Bitcoin Cash, Zcash, Bitcoin SV, Dash and Monero. All except Monero require ASICs for efficient mining. Only Bitcoin and Ethereum mining pools are found in the current top 10.
The full list with additional explanation is available in a slide presentation at OrionX.net/research.
“The traditional data center is becoming a supercomputer center,” said Shahin Khan, OrionX Partner and Analyst. “Corporate IT departments increasingly rely on versatile supercomputers as the engines of digital transformation. Green500, IO500, CryptoSuper500 and similar competitive lists point to the versatility of supercomputers and the need for supercomputing in Fortune 500.”
About OrionX
OrionX is a Silicon Valley firm with a new model that combines technology research, market execution, and customer engagement. More than 60 technology leaders in virtually every technology segment have trusted OrionX to help set new break-away strategies, ignite brands, and grow market share. Visit us at OrionX.net.
* Note: This effort is an analysis of the technologies and trends surrounding blockchain and cryptocurrencies. It is not, and must not be considered as, financial, investment, or legal advice.
References:
Overall: coinmarketcap.com, coinwarz.com, cryptoslate.com
BTC: btc.com
ETH: etherscan.io, btc.com
BCH: btc.com, cash.coin.dance
LTC, ZEC, XMR, DASH: miningpoolstats.stream
Cryptocurrency topics: orionx.net/blog
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.
Cryptocurrency mining pools have become larger. A new list, CryptoSuper500 tracks them according to economical value, computational power, and by extension, energy use.
If you are following my colleague Dr. Stephen Perrenod‘s blogs and reports, you know that he, and the rest of us at OrionX, aim to keep on top of business, technology, and policy issues surrounding this technology and take a realistic approach to what Cryptocurrencies can do, and under what circumstances their use is a wise choice.
The amount of energy that is used for crypto mining has been a source of controversy. It only applies to cryptocurrency mining via Proof of Work (PoW), where the “work” is computationally intensive. PoW mining is important since it represents the most effective consensus algorithm to maximize security in a decentralized manner. Mined cryptocurrencies continue to be in demand, with relatively high values. We expect this to continue despite significant ups and downs. We also expect efforts to create Proof of Useful Work (PoUW), but that is proving elusive at present, though I have not seen a proof that it cannot be done.
There are many other cryptocurrencies where the coin is not mined at all. Those tend to offer faster transactions and consume less energy. There are a few models here.
“Proof of Stake” (PoS) is the most prominent of what I call the “Proof of <insert costly thing>” approaches. Other well-known models include Byzantine Fault Tolerance (BFT) and Directed Acyclic Graphs (DAG). These algorithms do not mine and thus do not make use of specialized supercomputer resources. They are fine for many use cases and are produced more readily, but they tend to have lower security, and none are fully decentralized solutions.
So PoW mining uses substantial computing power, but it delivers something that is important and in demand. And if you look at other ways of achieving trust, it turns out that they are expensive in one way or another. So advocates believe crypto mining energy use compares favorably with the energy costs associated with traditional mining, refining, storage, and transport of precious metals.
So the list at Cryptocurrency Supercomputing, the large-scale cryptocurrency mining operations which are typically powered by accelerator technologies such as GPUs, FPGAs, or custom ASICs. Bitcoin is the most notable of such currencies. Scroll down for the full table.
Keeping track of these looks very similar to the TOP500 list of most powerful supercomputers. A list that has spawned variations that look at different use cases, for example, the Green500, Graph500, and IO500 lists. CryptoSuper500 was inspired by those lists.
We issued the following press release today to formally announce this list. Read it below or here and scroll down for the full table.
Bitcoin and BTC.com Top New CryptoSuper500 List
Inspired by the TOP500, biannual list of the most powerful cryptocurrency mining pools tracks compute power and energy useMENLO PARK, Calif. – Nov. 12, 2018 – PRLog — OrionX Research today announced that Bitcoin and BTC.com top the first release of the CryptoSuper500 list. The list recognizes cryptocurrency mining as a new form of supercomputing and tracks the top mining pools.
“CryptoSuper500 tracks supercomputers that are used for cryptocurrency mining, an intensive application that has become a driver of technology development and investment decisions globally,” said Dr. Stephen Perrenod, OrionX Partner and Analyst and the developer of the CryptoSuper500 list. “The growth of the cryptocurrency market has put the spotlight on emerging decentralized applications, the new ways in which they are funded, and the software stack on which they are built.” Cryptocurrency technologies include blockchain, consensus algorithms, digital wallets, and utility and security tokens.
The top 5 pools are BTC.com in China with projected annual economic value (AEV) of $694 million, followed by Antpool based in Hong Kong with an AEV of $582 million, SlushPool and ViaBTC, both in the USA with equal AEV of $444 million each, and BTC.Top of China with an AEV of $400M. The full list is attached below and is available with additional explanation in a slide presentation and white paper at OrionX.net/research.
Cryptocurrency mining via “Proof of Work” is important since it continues to represent the most effective consensus algorithm to maximize security in a decentralized manner. Cryptocurrency Supercomputing refers to large-scale cryptocurrency mining operations which are typically powered by accelerator technologies such as GPUs, FPGAs, or custom ASICs. Bitcoin is the most notable of such currencies.
Other protocols and methods including premining, Proof of Stake, and Byzantine Fault Tolerance, eschew mining and thus do not make use of specialized supercomputer resources. Such tokens are produced more readily, tend to have lower economic value and security, and none are fully decentralized solutions.
At the same time, the amount of energy that is used for crypto mining has been a source of controversy even though it compares favorably with the energy costs associated with traditional mining, refining, storage, and transport of precious metals.
“TOP500, Green500, IO500, HPCG, and now CryptoSuper500 all point to the growing versatility of supercomputers,”
said Shahin Khan, OrionX Partner and Analyst. “It is time to more explicitly recognize that supercomputers are, and will be, central to important new business applications. I expect we will see similar competitive lists for AI, Quantum Computing, and public Cloud providers.” The TOP500 list of the world’s most powerful supercomputers is eagerly awaited twice a year, and the top position has been a matter of national pride and intense competition among vendors.
CryptoSuper500 is the first compilation of such a list across multiple cryptocurrency types which use different mining algorithms and blockchain protocols. CryptoSuper500 uses the economic value of the mined coins produced to make comparisons across coin types and mining pools when rank ordering the list. The first release of the list includes the Top 30 mining pools and their daily economic production, considering which crypto of the top 5 was produced.
About OrionX
OrionX is a Silicon Valley firm with a new model that combines technology research, market execution, and customer engagement. More than 60 technology leaders in virtually every technology segment have trusted OrionX to help set new break-away strategies, ignite brands, and grow market share. Visit us at OrionX.net.* Note: This effort is an analysis of the technologies and trends surrounding blockchain and cryptocurrencies. It is not, and must not be considered as, financial, investment, or legal advice.
References:
Overall: coinmarketcap.com, coinwarz.com
BTC: btc.com, investoon.com
ETH: etherscan.io
BCH: btc.com, cash.coin.dance
LTC, XMR, DASH: miningpoolstats.stream
Cryptocurrency topics: orionx.net/blog
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.
National Security considerations are often intimately intertwined with the adoption of new technologies.
Last year Tokyo hosted a meeting of the International Standards Organization, including a session on blockchain technology to examine ideas around standards for blockchain and distributed ledgers.
A member of the Russian delegation, who is part of their intelligence apparatus at the FSB, apparently said “the Internet belonged to the US, the Blockchain will belong to Russia.” In fact three of the four Russian delegates were FSB agents! By contrast, Chinese attendees were from the Finance Ministry, and American attendees were representing major technology companies, reportedly IBM and Microsoft among others.
Let’s unpack this a bit. The Internet grew out of a US military funded program, Arpanet, and the US has been the dominant player in Internet technology due to the strength of its research community and its technology companies in particular.
As we wrote in our most recent blog (https://orionx.net/2018/05/is-blockchain-the-key-to-web-3-0/), blockchain has the potential to significantly impact the Internet’s development, as a key Web 3.0 technology.
Blockchain and the first cryptocurrency, Bitcoin, were developed by an unknown person or persons, with pen name Satoshi Nakamoto. Based on email timestamps, the location may have been New York or London, so American or British citizenship for Bitcoin’s inventor seem likely, but that is speculation.
More to the point, the US is the center of blockchain funding and development activity, while China in particular has been playing a major role in mining and cryptocurrency development.
There are many Russian and Eastern European developers and ICO promoters in the community as well. The Baltic nations bordering Russia and the Russian diaspora community have been particularly active.
The second most valuable cryptocurrency after Bitcoin is Ethereum, which was invented by a Russian-Canadian, Vitalik Buterin. Buterin famously met with Russian President Vladimir Butin in 2017. Putin is himself of course a former intelligence agent.
The Russians reportedly want to influence the cryptographic standards around blockchain. This immediately raises fears of a backdoor accessible to Russian intelligence. Russia is also considering the idea of a cryptocurrency as a way to get around sanctions imposed by the American and European governments.
The Russian government has a number of blockchain projects. The government-run Sberbank had initial implementation of a document storage blockchain late last year. There is draft regulation around cryptocurrency working its way through the Russian parliament. President Putin has said that Russia cannot afford to fall behind in blockchain technology.
Given the broad array of applications being developed for cryptocurrencies, including money transfer, asset registration, identity, voting, data security, and supply chain management among others, national governments have critical interests in the technology.
China has been cracking down on ICOs and mining, but it is clear they think blockchain is important and they want to be in control. Most of their government concerns and interest appear to be centered around the potential in finance, such as examining the possibility of a national cryptocurrency (cryptoYuan).
China would like to wriggle free from the dollar standard that dominates trade and their currency reserves. They have joined the SDR (foreign reserve assets of the IMF) and have been building their stocks of gold as two alternatives to the dollar.
China’s biggest international initiative is around a new ‘Silk Road’, the One Belt, One Road initiative for infrastructure development across EurAsia and into the Middle East and Africa. One could imagine a trading currency in conjunction with this, a “SilkRoadCoin”. In fact, the government-run Belt and Development Center has just announced an agreement with Matrix AI as blockchain partner. Matrix AI is developing a blockchain that will support AI-based consensus mechanisms and intelligent contracts.
China’s One Belt One Road Initiative actually has six land corridors and a maritime corridor.
(Image credit: CC 4.0, author: Lommes)
The American military is taking interest in blockchain technology. DARPA believes that blockchain may be useful as a cybersecurity shield. The US Navy has a manufacturing related application around the concept of Digital Thread for secure registration of data across the supply chain.
In fact the latest National Defense Authorization Act requires the Pentagon to assess the potential of blockchain for military deployment and to report to Congress their findings, beginning this month for an initial report.
What is clear, is that blockchain and distributed ledger technology have the potential to be of major significance in national security and development for the world’s leading nations.
A range of efforts are underway by government, industry, and academia to understand blockchain technology and cryptocurrencies, to enhance the technology, and plan for the future. In that context, we see potential in the technology to impact many facets of society and global dynamics. That potential is now sufficiently developed for us to advocate a more visible presence by government agencies in helping shape the policy and academic research.
We encourage the US government to increase engagement with blockchain and distributed ledger technology. This can include funding research in universities, pilot projects with industry across various government agencies including the military and intelligence communities, the Federal Reserve, and the Department of Energy, NOAA and NASA, in particular.
Also the federal government should pursue standards development under the auspices of the NIST and together with ISO. Individual state governments are also promising laboratories for projects around identity, voting, and title registration.
Information has always been key to warfare. But there is little doubt that warfare is increasingly moving toward a battlefield within the information sphere itself. These are wars directed against the civilian population; these are wars for peoples’ minds. Blockchain technologies could play a significant role in these present and future battles, both defensively and offensively.
America was founded and grew rapidly largely in the context of the Industrial Revolution. The Information Age provides a similar opportunity and responsibility to set the course for the next century and beyond. As before, getting it right will not only assure the country’s continued success and leadership. It also arms the nation to solve problems that affect all of humanity.
References :
DARPA https://www.google.co.th/amp/s/cointelegraph.com/news/pentagon-thinks-blockchain-technology-can-be-used-as-cybersecurity-shield/amp
US Navy http://www.secnav.navy.mil/innovation/Pages/2017/06/BlockChain.aspx
2018 National Defense Authorization Act https://www.realcleardefense.com/articles/2018/05/03/could_americas_cyber_competitors_use_blockchain_for_their_defense_113400.html
NIST https://csrc.nist.gov/publications/detail/nistir/8202/draft
http://www.scmp.com/business/china-business/article/2136188/beijing-signals-it-wants-become-front-runner-blockchain
https://www.cbinsights.com/research/future-of-information-warfare/
Stephen Perrenod has lived and worked in Asia, the US, and Europe and possesses business experience across all major geographies in the Asia-Pacific region. He specializes in corporate strategy for market expansion, and cryptocurrency/blockchain on a deep foundation of high performance computing (HPC), cloud computing and big data. He is a prolific blogger and author of a book on cosmology.
I am very pleased and excited to announce the release of the newest paper in our Constellation™ series about blockchain/cryptocurrency technologies: “Cryptocurrency Outlook: More Chasm Crossing.”
This is a comprehensive look at this emerging and important technology, written by my colleague Dr. Stephen Perrenod who has been working with cryptocurrencies for several years and has focused his research on this topic for more than a year,
To learn more, read the press release below and view the paper in full at OrionX.com/Blockchain.
We are issuing the following press release to formally announce this paper:
A comprehensive look at what lies ahead in the cryptocurrency economy
Menlo Park, CA, April 2, 2018: OrionX, a Silicon Valley-based technology research, market execution, and customer engagement firm, today released a new research paper in its Constellation™ series. Written by OrionX partner, Dr. Stephen Perrenod, “Cryptocurrency Outlook: More Chasm Crossing” provides an in-depth view of the current cryptocurrency economy and an analysis of what might be expected in the next 9-15 months*.
“Blockchain and cryptocurrency are at the forefront of online conversation,” said Dr. Stephen Perrenod. “Everyone from novice investors to seasoned economists is working to develop a new level of understanding of ‘money in the internet’ as cryptocurrency value remains volatile but seems to show staying power.” Perrenod’s research breaks down the specifics of how this technology works and what the future may hold in terms that are understandable to anyone seeking an education on what he calls “Money 3.0.”
The paper provides a detailed look at how cryptocurrency is evolving, with special attention paid to the key points below:
- While many criticisms of Money 3.0 focus on technology issues like scaling, those issues are being addressed at a rapid pace.
- Detailed view of the top five crypto “coins,” including key metrics on how they have performed in the past as well as parameters that can impact the future.
- Predictions about how traditional institutions—banks and governments—may affect cryptocurrency markets in the coming months.
- Amazon and PayPal changed the way we spend money, now cryptocurrencies are changing the way it is created and distributed.
In addition, the paper makes a number of predictions as conclusions to discussions of the relevant topics. Examples include:
- Both the number of Ethereum addresses and the Ethereum market cap will exceed those for Bitcoin by the end of 2018.
- Increasing regulation of ICOs and exchanges will open the taps for Wall Street investment.
- The first unicorn ICO will occur this year.
- The era of Bitcoin forks is largely behind us.
- Dapps, decentralized apps, will blossom wildly.
- We will see multiple social media blockchain unicorns. Facebook could react by implementing its own token, in fact they would be remiss not to.
- Existing centralized exchanges will provide the facility for direct exchange from an individual’s wallet, mimicking the peer-to-peer experience of distributed exchanges.
To view the paper in full, please visit OrionX.net/blockchain
“Guided by its diverse technology clients, OrionX conducts research in advanced topics such as Internet of Things (IoT), Artificial Intelligence (AI), Blockchain/Cryptocurrencies, and Quantum Computing,” said Shahin Khan, founding partner at OrionX. The group’s technology foundation is in High Performance Computing (HPC), Data Center technologies, and Cloud Computing, which inform its general perspective.
OrionX also announced that it will present at the Advanced Scale Forum 2018 conference to be held May 6-8, 2018 in Austin, Texas. OrionX founding partner, Shahin Khan will present “Blockchain, Crypotocurrencies, Smart Contracts – Technologies You Can’t Ignore” in collaboration with Ryan Quick, principal at Providentia Worldwide, an advanced technology consultancy. Mr. Quick previously led PayPal’s Advanced Technology Group. The session will cover the essence of blockchain technologies, focusing on their impacts to other key trends (such as IoT, AI, and quantum computing), identifying their realities and promises, and exploring why blockchain may or may not deliver.
This talk follows Mr. Khan’s futuristic presentation (available from slideshare.com) on “Critical Technologies of the Future” at the HPC-AI Advisory Council conference held at Stanford University in Feb-2018. A video of the talk is available from InsideHPC.com.
* This paper is an analysis of the technologies and trends surrounding blockchain and cryptocurrencies. It is not, and must not be considered as, financial, investment, or legal advice.
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.
Warren Buffett is the most famous name in stock investing, the second richest person in the world, and a leading expert in valuing companies and securities. He also is a big investor in banks, including Wells Fargo, Goldman Sachs, and Bank of America. So he has a lot of vested interest in the current monetary and financial system.
He recently said that bitcoin is in a bubble, which may well be true. He also said “You can’t value bitcoin because it’s not a value-producing asset”.
And yet, as of today, bitcoin has a market cap which is over $120 billion, substantially more than Buffet’s net worth. Bitcoin can be valued, but not like a security or a company, which is what Buffet does so well.
As we have written elsewhere, bitcoin is a currency, money, a cryptocurrency. Some will say it is a digital asset, like a digital gold. We say it is Money 3.0 (fiat is Money 2.0, gold and silver coins were Money 1.0).
Do these have value? Although none has intrinsic value and each costs a nominal amount to make the one on the right is worth quite a bit more than the one in the center, and the one on the left, more still.
Yes, money is designed as a store of value, a medium of exchange, a unit of account. Cryptocurrencies have the same design goals.
Bitcoin is a unit of account in a secure, distributed ledger. It has been a much better store of value than fiat currency in recent years.
In fact in 2014, Buffet said “It’s a mirage”. And yet, the value of a single bitcoin has increased 20 times since late October, 2014 to about $5800 today.
Money is valued by what you can exchange it for in the economy, by its utility. It is valued against other currencies. Bitcoin can be valued the same way, and thus by the vitality of the bitcoin economy.
Within the cryptocurrency world, bitcoin is the reference benchmark, just as the US dollar is globally. Bitcoin gets valued every single day, every minute of the day, against all major fiat currencies and against hundreds of cryptocurrencies. Like those currencies it has an economy and a turnover (or velocity) of the currency.
One thing we don’t normally think of with respect to cryptocurrencies is interest, or dividends. It is possible to lend out bitcoin for interest, as one can do with dollars, euros, or even gold. But bitcoin has effectively thrown off two special dividends this year, in the form of forks of the bitcoin blockchain. These are known as Bitcoin Cash (forked in August) and Bitcoin Gold (forked in October). Collectively, those two are worth close to $700, representing over 9% dividend rate to date during 2017, based on the current bitcoin price.
And for someone who bought at the beginning of the year, when bitcoin was under $1000, the dividend is around 70%. Not shabby, and a very reasonable reward for accepting the volatility.
Bitcoin is a technology for internet money, network money that is independent of any government, and it can be hard to fathom for the newcomer. Buffett has always said he does not understand technology. He was late getting into Apple, for example. He has not examined the technology and potential of bitcoin and other cryptocurrencies sufficiently to have an informed opinion.
Bitcoin’s future value? It all depends on how the economy around bitcoin develops, but it will be quite an adventure. And bitcoin to bagels it will increase in value over the next several years.
Stephen Perrenod has lived and worked in Asia, the US, and Europe and possesses business experience across all major geographies in the Asia-Pacific region. He specializes in corporate strategy for market expansion, and cryptocurrency/blockchain on a deep foundation of high performance computing (HPC), cloud computing and big data. He is a prolific blogger and author of a book on cosmology.
Why are cryptocurrencies like Bitcoin and Ethereum Money 3.0? And what are Money 1.0 and Money 2.0?
Recently, Jamie Dimon called Bitcoin a ‘fraud’. This coming from the CEO of JP Morgan, the bank that has been fined more than any other, save one, for financial crimes since the Great Recession of 2008. His statement reeks of hypocrisy since JPM is a member of the Enterprise Ethereum Alliance, his traders have been trading Bitcoin related ETN securities, and his firm has applied for patents using blockchain technology.
By the way the Enterprise Ethereum Alliance has well over 100 members including Microsoft and Mastercard. Serious players understand that cryptocurrencies are a big deal. The market cap of all cryptocurrencies is currently in the neighborhood of $150 billion, around 2/3 the market cap of Visa. And this has all happened in only eight years’ time.
It is amazing how few people can give a definition, other than pulling out a bill from their wallet, or referring to the numbers in their checking account statement. And how does money get created in our modern economy? Very few actually understand the process. Most people say government creates it. Governments can, and do, but most money is not created directly by the government. What the government does is validate money, they define a single type of money for their nation. They print currency, but most money today is digital, residing in bank balance sheets, and most money creation occurs as banks issue new loans.
Throughout history there have been many forms of money, but two forms have dominated. The first form, Money 1.0, was the dominant form for millennia. It was coins made of precious metals, in particular gold and silver, and ‘base’ metals such as copper and bronze. According to the St. Louis Federal Reserve, money must have six properties: durability, portability, divisibility, uniformity, limited supply, and acceptability.
They sound a bit like goldbugs when they write it that way. These are all useful attributes of the thing that is used for money, be it gold or paper. But it doesn’t quite get to the most essential three properties of money. It must serve as:
Money is whatever can be used as a socially agreed upon unit of account and medium of exchange. It also should retain its value, not depreciate quickly, so that it can be used next month and next year as well. Notice I say socially. Societies agree upon what is used as money, and nation states in recent centuries have taken the lead in that definition. In order to be conveniently exchanged, then the six properties above come into play. Durability and limited supply allow the retention of value. Portability and divisibility make it easier to exchange. Uniformity makes it a useful unit of account, as does acceptability.
We all have to more or less agree on what the accounting unit is. That is actually the starting point for money, agreeing on the standard measure. The government can decree the accounting unit, and can demand taxes be paid in that unit. That is government fiat, and can apply for either coined money of precious or base metals (Money 1.0) or paper money (Money 2.0).
Image: Roman gold Solidus coin. York Museums Trust. CC-BY-SA 4.0
The US dollar was originally defined to contain a certain weight of silver, and aligned to the Spanish dollar (originally Austrian thaler) or ‘pieces of eight’ that was widely used in New World trade. The US dollar has also been defined against gold, with an official act in 1900 following nearly 3 decades of defacto gold standard following the Civil War. Of course the gold standard is now entirely gone after being discarded in two phases, under Roosevelt in 1933 and Nixon in 1971. The remnants of the bimetallic standard of the late 19th century remain in present-day dimes and quarters that used to contain silver even until 1965, retain the color, but have been entirely debased.
No nation remains on a Money 1.0 standard of precious metals, all have moved to Money 2.0, fiat paper money. If they did they would lose their gold, and they prefer to melt it into bars and store it in central bank vaults as a reserve. So as Warren Buffet says, we dig it up in mines, melt it down into bars, and bury it again in vaults.
With paper money, there must be fiat, as nobody wants pieces of paper that have no value. The days of gold certificates and silver certificates as circulating currency are long gone, although I remember silver certificate dollar bills from my youth. The value comes from the legal tender requirements that the paper be accepted by businesses, be used for taxes, and from the government’s printing process to make counterfeiting difficult plus the government’s overall management of the money supply (usually through interest rate policies) to limit loss of value due to inflation.
The technology of high quality paper engraving, augmented with serial numbers, threads and holograms, and the technology of central banks, allow fiat money to work. The vast majority of nations have central banks to lend to the commercial banks in times of crisis and to manage the banking system and money supply indirectly.
So those are Money 1.0 and Money 2.0. In summary:
Money 1.0 – Public or private, asset-based, intrinsic value, coins or bars of precious metal
Image of $2 Federal Reserve Note, Bureau of Engraving and Printing, U.S. Dept. of Treasury
Money 2.0 – Public and sovereign, debt-based, no intrinsic value, paper and digital.
Most Money 2.0 is digital, with the circulating currency representing a small percentage. Money mostly comes into circulation not through the printing press, but when banks make new loans. If a bank creates an auto loan, it credits the checking account of its customer digitally. Banks are allowed to make new loans within the limits of their central bank authority determined reserves and equity capital requirements.
Note as an aside that Money 1.0 and Money 2.0 can coexist. We mostly have Money 2.0 in the United States, but there was a small amount of silver coinage money circulating alongside up until the 1960s. This is an important principle, since we are beginning to see the coexistence of Money 2.0 and Money 3.0.
Cryptocurrencies are purely digital, whereas Money 2.0, fiat and debt-based money, is mostly digital.
Why Money 3.0? Technologists and advocates of non-fiat money were concerned about the risks of centralized monetary systems dominated by central banks and by money center banks engaged in fraudulent activities around mortgages and other lending with derivatives including CDOs, CDSs and more. The corrupt system lead to the Great Recession of 2008. Everyone in the society suffered, but the banks were bailed out by enormous government loans.
There were more than 50 attempts at creating a digital crypotcurrency prior to the year 2000. None succeeded. One was gold-based and known as e-gold. It was shut down in 2009 by the US government, because it ran afoul of stricter money laundering regulations. It was also subject to repeated thefts of accounts from Russian and other criminal hackers.
A successful non-fiat cryptocurrency must provide a single secure ledger of entries to protect against counterfeiting and double spending. It must have a method of commiting a single instance of a transaction to this secure ledger that is publicly shared, and is known as the blockchain. It must have a built-in automated “central banking” function that determines the money supply.
Satoshi Nakamoto’s brilliance was to combine a number of existing ideas around public/private key cryptography, distributed ledgers, and a mining algorithm with “proof of work” that rewarded miners for solving a difficult cryptographic hash problem. Transactions are signed with private keys. All bitcoins reside in the distributed ledger. The owner has a wallet with the key that allows them to transfer bitcoin in arbitrary amounts to someone else and thus confers ownership.
The supply is limited with a maximum at 21 million bitcoins that will not be reached until well into the 22nd century. New bitcoin comes into existence in conjunction with the mining of blocks of transactions. The successful miner is rewarded with an allocation of new coins, presently 12.5 coins per block of approximately 2000 transactions. So here we have the central banking function and a digital minting or mining process for the ‘coins’ which are really just ledger entries.
We describe this Nakamoto consensus algorithm and the mining process in more detail at orionx.net/podcast.
Now we have not just Bitcoin, but Ethereum, Bitcoin Cash (which is a recent fork of Bitcoin with large block size), Ripple, Litecoin and hundreds more cryptocurrencies. We have new coins being created rapidly in conjunction with new applications and ICOs – initial coin offerings.
The largest of these, those with market caps in the billions of dollars, meet the three requirements for money. Unit of account. Medium of Exchange. Store of value. Their limitations at present relate to the latter two attributes. They are accepted as medium of exchange in some environments, but relatively few compared to existing fiat currencies. And as a result of that their value is less stable and determined more by investment and speculative demand. Their ultimate value will be determined by the cryptocurrency economy as uses cases, applications, and acceptance grow.
They are child currencies, developing and growing, but far from the maturity of an existing national fiat currency. The value should continue to grow for the long term, however since transaction volumes are increasing very rapidly.
So now we have in the world:
Money 3.0 – Private and globally distributed, asset-based, digital only.
Money 3.0 holds much promise. It can remove a lot of cost and friction from the financial system. Trying sending a check or ACH transfer to your sister and having the transaction complete on the weekend. Send her some bitcoin? She will get it even on Sunday at 3 am around an hour or so after you send it. Bitcoin is 24 by 7 by 365. And with very limited fees within the Bitcoin economy. Most of the cost is in moving Bitcoin to fiat or vice versa.
It is not based on debt, so does not have the instability of debt or counterparty risk. The only real risk is security, which holds as well for your banking balances. The other risk is to the value as governments and politicians feel threatened. But at the end of the day, they can only regulate it, but not eliminate it. The technology is too widely available to anyone.
Money 3.0 is not poised to replace Money 2.0 anytime soon, although in a number of ways it is superior. They will coexist. At some point a small country will convert their currency to Money 3.0, by building a blockchain-based peso or some such. A number of central banks, large and small, are already studying this issue.
Many have talked about global currencies in the past. The US dollar has global impact for trade and the price of key commodities, but you have to exchange it when you cross borders. The Euro has been a boon for commerce, trade, and travel in many countries within Europe. Gold historically had a global role but was difficult to move and verify as to weight and purity.
Bitcoin has no weight and purity issues. It transcends borders. It, Ether, and the other cryptocurrencies are indeed the first global currencies.
Stephen Perrenod has lived and worked in Asia, the US, and Europe and possesses business experience across all major geographies in the Asia-Pacific region. He specializes in corporate strategy for market expansion, and cryptocurrency/blockchain on a deep foundation of high performance computing (HPC), cloud computing and big data. He is a prolific blogger and author of a book on cosmology.
There are more Things than anything else! And they’re going online in droves, adding capability but also vulnerability, and making the Secure Internet of Things (IoT), the super set of the technology trends. Secure IoT is the place with the most interdisciplinary, and therefore, the most difficult challenges. If you believe, as you should, that no-compromise cybersecurity is table stakes, then IoT means “secure IoT” and IoT by itself becomes the ultimate hashtag soup. Here is why.
IoT and security require an end-to-end mindset. Today’s solutions are fragmented. Integrating them is hard and requires blending of many technologies and algorithms. It requires a team with a very broad set of expertise who have managed to understand each other and employ each other’s strengths.
Things come in many shapes and sizes. We have to make a distinction between Small-t things, sensors and micro devices, Big-T Things, which often form meta things or Systems, or All-caps THINGS, meta systems and the so-called Distributed Autonomous Organizations (DAOs). Managing them in a coherent manner requires an expertise rubber band that has to stretch across from the microscopic to the astronomic:
Security is not just a technology and practice but a non-negotiable requirement that cuts across all aspect of tech. Small-t things require a whole a new approach to device and data security all the way up the supply chain, a difficult technological and operational task an area of intense innovation, for example by blending AI, IoT, and Cybersecurity disciplines or pursuing entirely new approaches. Big-T Things remain hard to secure but have more built-in resources and processing capability, making them eligible to be handled like existing IT systems.
Things, scattered around, are exposed to physical tampering so it is important to physically secure them and detect any unauthorized physical access or near-field or far-field monitoring. As is often said, “security is security” so physical security and cybersecurity must be viewed and implemented as one thing: all under the umbrella of risk and threat management and #sustainability.
Many Things move, either by themselves (like #drones or robots) or carried by other things (pick a form of transportation) or by people (personal device or #wearables). Mobility issues from initial on-boarding to authenticated secure transmission to data streaming and data semantics to over-the-air (OTA) updates all become critical in IoT.
Distributed Things that generate lots of data, require secure authenticated communications, are possibly subject to compliance regulations, may need digital rights management (#DRM), or need verifiable delivery… are the kind of use case that Blockchain technology can address very well. IoT applications are often natural candidates. (see also the OrionX Download podcast series on Blockchain.)
Connected devices generate data. Making sense of that data requires Big Data practices and technologies: data lakes, data science, storage, search, apps, etc.
When you have so much data, invariably you will find an opportunity for AI (see the OrionX/ai page for a repository of reports), either via #MachineLearning which uses the data, or #ExpertSystems for policy management which decides what to do about the insight that the data generates. AI framewroks, especially the #DeepLearning variety, are #HPC oriented and require knowledge of HPC systems and algorithms.
There will come a time when another emerging trend, #QuantumComputing, will be useful for AI and cybersecurity.
If Things are smart enough and can move, they become robots. Distributed Autonomous Objects (DAOs) use and generate data and require a host of security, processing, control, policy, etc.
All of this processing happens in the cloud somewhere, so all cloud technologies come into play. The required expertise will cover the gamut: app development, streaming data, dev-ops, elasticity, service level assurance, API management, microservices, data storage, multi-cloud reliability, etc.
IoT provides new kinds of information extraction, manipulation, and control. Just the AI and Robotics components are enough to pose a challenge for existing legal systems. Ultimately, it requires an ethics framework that can be used to create a proper legal system all the while ensuring widespread communication so organizational structures and culture can keep up with it.
Not every IoT deployment will need all of the above, but as you try to stitch together a strategy for your IoT project, it is imperative to do so with a big picture in mind. IoT is the essence of #DigitalTransformation and touches all aspects of your organization. It’s quite a challenge to make it seamless, and you’ll need specialists, generalists, and not just analysts but also “synthesists”.
In the OrionX 2016 Technology Issues and Predictions blog, we said “If you missed the boat on cloud, you can’t miss it on IoT too”, and “IoT is where Big Data Analytics, Cognitive Computing, and Machine Learning come together for entirely new ways of managing business processes.” Later, in February of 2016, my colleague Stephen Perrenod’s blog IoT: The Ultimate Convergence, further set the scene. Today, we see those predictions have become reality while IoT is poised for even more convergence.
We will cover these topics in future episodes of The OrionX Download podcast, identifying the salient points. As usual, we will go one or two levels below the surface to understand not just what these technologies do, but how they do it and how they came about. That will in turn, help put new developments in perspective and explain why they may or may not address a pain point.
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.
We are very pleased to announce the completion of our “epic” AI survey. Here is the press release that we issued this morning:
ORIONX RESEARCH ANNOUNCES SURVEY RESULTS REVEALING CUSTOMERS’ ADOPTION OF ARTIFICIAL INTELLIGENCE (AI), MACHINE LEARNING (ML) AND DEEP LEARNING (DL)
MENLO PARK, Calif., August 30, 2017 – OrionX Research today announced the availability of survey results of more than 300 North American companies representing 13 industries on what they are currently doing and planning to do with Artificial Intelligence (AI), Machine Learning (ML) and Deep Learning (DL) technologies.
Considered one of the most comprehensive surveys to date on AI/ML/DL, the survey covers 144 questions/data points, presented in over 70 charts and graphs. It explores topics ranging from key drivers for AI project implementation to customer buying behavior to current and future AI budget spend to the most popular AI models.
Study highlights include:
- Top challenges businesses are facing for their AI/ML/DL projects
- Current and planned budgets for AI/ML/DL projects (broken down by hardware, software, services, cloud, consulting, education/training, staffing and more)
- Who the decision makers are and which departments are driving AI purchasing decisions
- Average AI project data size (TB) and the application areas that AI projects are being used for
- User ranking of over 20 AI Frameworks with TensorFlow dominating at 52%
- AI hardware usage (now and in the future), including most used hardware accelerators with NVIDIA in the lead at 82%
- Customers’ key selection criteria and requirements when evaluating AI vendors
“The OrionX AI survey is the first to assess customer sentiment across a wide range of AI adoption issues,” said Shahin Khan, founding partner at OrionX. “Among many insights, it shows how customer selection criteria is influenced by parameters such as vendor market presence, product roadmap, current product features, ease of adoption, ease of use, interoperability, security, performance, and robustness.” This comprehensive AI study will help customers better understand the AI/ML/DL landscape in order to set strategy and guide future investments.
To purchase the report and learn how OrionX research analysts can help companies navigate the multi-faceted and rapidly evolving AI/ML/DL marketplace, visit www.orionx.net/ai
Sample Chart from the OrionX Research AI Survey
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.
Data loss is a subject whose time has come! Digital Transformation is upon us and with it the beginnings of a vibrant data economy. Extracting, manipulating, and making sense of the information content from physical assets and business processes is the essence of this new era. With it, come two growing problems: downtime, and data loss.
Downtime is interruption, the opposite of business continuity. In a digital enterprise, more than ever, business continuity depends on high availability (HA) and disaster recovery (DR). Both of those depend on fast data replication.
Downtime has been studied for decades. Preventing it is better understood, but it continues to happen from time to time. Sometimes it’s kept quiet and sometimes it’s spectacularly public. But it always causes damage. When a major airline experienced IT failures recently, it not only made the headlines but it cost the company millions of dollars in lost revenue, and perhaps also in damaged brand image. Search online for “IT outage” and you will see sobering stories. Downtime is serious business.
Digital organizations are at the nexus of massive data flowing in and out of them. As they become more common, the value of an organization’s data assets become correspondingly more significant. As a result, one aspect of downtime that is getting more attention is the value of the data that is lost or not captured.
What is the cost to your organization if you lose your payroll, or your customer transactions for a day’s business? These painful types of data loss are harder to quantify and more importantly, ones your company doesn’t want to encounter.
In our recent research paper, The Cost of Data Loss and How to Avoid It, my colleague Dan Olds examines the drivers and cost categories of data loss. He has also looked at what solutions are available to prevent data loss, and highlights a particularly effective one that is based on Open Source technologies and has been downloaded an incredible 1.7 million times.
Browse through all of our free Constellation papers, or go directly to the full Data Loss paper to learn more.
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.
Here at OrionX.net, our research agenda is driven by the latest developments in technology. We are also fortunate to work with many tech leaders in nearly every part of the “stack”, from chips to apps, who are driving the development of such technologies.
In recent years, we have had projects in Cryptocurrencies, IoT, AI, Cybersecurity, HPC, Cloud, Data Center, … and often a combination of them. Doing this for several years has given us a unique perspective. Our clients see value in our work since it helps them connect the dots better, impacts their investment decisions and the options they consider, and assists them in communicating their vision and strategies more effectively.
We wanted to bring that unique perspective and insight directly to you. “Simplifying the big ideas in technology” is how we’d like to think of it.
Thus was born our first podcast/slidecast series: The OrionX Download™
The OrionX Download is both a video slidecast (visuals but no talking heads) and an audio podcast in case you’d like to listen to it.
Every two weeks, co-hosts Dan Olds and Shahin Khan, and other OrionX analysts, discuss some of the latest and most important advances in technology. If our research has a specially interesting finding, we’ll invite guests to probe the subject and add their take.
Please give it a try and let us know what we can do better or if you have a specific question or topic that you’d like us to cover.
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.
Just as Intel, the king of CPUs and the very bloodstream of computing announced that it is ending its Intel Developer Forum (IDF) annual event, this week in San Jose, NVIDIA, the king of GPUs and the fuel of Artificial Intelligence is holding its biggest GPU Technology Conference (GTC) annual event yet. Coincidence? Hardly.
With something north of 95 per cent market share in laptops, desktops, and servers, Intel-the-company is far from even looking weak. Indeed, it is systematically adding to its strengths with a strong x86 roadmap, indigenous GPUs of its own, acquisition of budding AI chip vendors, pushing on storage-class memory, and advanced interconnects.
But a revolution is nevertheless afoot. The end of CPU-dominated computing is upon us. Get ready to turn the page.
Digitization means lots of data, and making sense of lots of data increasingly looks like either an AI problem or an HPC problem (which are similar in many ways, see “Is AI the Future of HPC?”). Either way, it includes what we call High Density Processing: GPUs, FPGAs, vector processors, network processing, or other, new types of accelerators.
GPUs made deep neural networks practical, and it turns out that after decades of slow progress in AI, such Deep Learning algorithms were the missing ingredient to make AI effective across a range of problems.
NVIDIA was there to greet this turn of events and has ridden it to strong leadership of a critical new wave. It’s done all the right things and executed well in its usual competent manner.
A fast-growing lucrative market means competition, of course, and a raft of new AI chips is around the corner.
Intel already acquired Nervana and Movidius. Google has its TPU, and IBM its neuromorphic chip, TrueNorth. Other AI chip efforts include Mobileye (the company is being bought by Intel), Graphcore, BrainChip, TeraDeep, KnuEdge, Wave Computing, and Horizon Robotics. In addition, there are several well-publicized and respected projects like NeuRAM3, P-Neuro, SpiNNaker, Eyeriss, and krtkl going after different parts of the market.
One thing that distinguishes Nvidia is how it addresses several markets with pretty much a single underlying platform. From automobiles to laptops to desktop to gaming to HPC to AI, the company manages to increase its Total Available Market (TAM) with minimal duplication of effort. It remains to be seen whether competitive chips that are super optimized for just one market can get enough traction to pose a serious threat.
The world of computing is changing in profound ways and chips are once again an area of intense innovation. The Silicon in Silicon Valley is re-asserting itself in a most welcome manner.
________
Note: A variation of this article was originally published in The Register, “The rise of AI marks an end to CPU dominated computing”.
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.
A variation of this article was originally published in The Register, “Wow, what an incredible 12 months: 2017’s data center year in review”, Feb 6, 2017.
The data center market is hot, especially now that we are getting a raft of new stuff, from promising non-Intel chips and system architectures to power and cooling optimizations to new applications in Analytics, IoT, and Artificial Intelligence.
Here is my Top-10 Data Center Predictions for 2017:
Data centers are information factories with lots of components and moving parts. There was a time when companies started becoming much more complex, which fueled the massive enterprise resource planning market. Managing everything in the data center is in a similar place. To automate, monitor, troubleshoot, plan, optimize, cost-contain, report, etc, is a giant task, and it is good to see new apps in this area.
Data center infrastructure management provides visibility into and helps control IT. Once it is deployed, you’ll wonder how you did without it. Some day, it will be one cohesive thing, but for now, because it’s such a big task, there will be several companies addressing different parts of it.
Cloud is the big wave, of course, and almost anything that touches it is on the right side of history. So, private and hybrid clouds will grow nicely and they will even temper the growth of public clouds. But the growth of public clouds will continue to impress, despite increasing recognition that they are not the cheapest option, especially for the mid-size users.
AWS will lead again, capturing most new apps. However, Azure will grow faster, on the strength of both landing new apps and also bringing along existing apps where Microsoft maintains a significant footprint.
Moving Exchange, Office, and other apps to the cloud, in addition to operating lots of regional data centers and having lots of local feet on the ground will help.
Some of the same dynamics will help Oracle show a strong hand and get close to Google and IBM. And large telcos will stay very much in the game. Smaller players will persevere and even grow, but they will also start to realize that public clouds are their supplier or partner, not their competition! It will be cheaper for them to OEM services from bigger players, or offer joint services, than to build and maintain their own public cloud.
Just as the hottest thing in enterprise becomes Big Data, we find that the most expensive part of computing is moving all that data around. Naturally, we start seeing what OrionX calls “In-Situ Processing” (see page 4 of the report in the link): instead of data going to compute, compute would go to data, processing it locally wherever data happens to be.
But as the gap between CPU speed and storage speed separates apps and data, memory becomes the bottleneck. In comes storage class memory (mostly flash, with a nod to other promising technologies), getting larger, faster and cheaper. So, we will see examples of apps using a Great Wall of persistent memory, built by hardware and software solutions that bridge the size/speed/cost gap between traditional storage and DRAM. Eventually, we expect programming languages to naturally support byte-addressable persistent memory.
Vendors already configure and sell racks, but they often populate them with servers that are designed as if they’d be used stand-alone. Vendors with rack-level thinking have been doing better because designing the rack vs the single node lets them add value to the rack while removing unneeded value from server nodes.
So server vendors will start thinking of a rack, not a single-node, as the system they design and sell. Intel’s Rack Scale Architecture has been on the right track, a real competitive advantage, and an indication of how traditional server vendors must adapt. The server rack will become the next level of integration and what a “typical system” will look like. Going forward, multi-rack systems are where server vendors have a shot at adding real value. HPC vendors have long been there.
Traditional enterprise apps – the bulk of what runs on servers – will show that they have access to enough compute capacity already. Most of that work is transactional, so their growth is correlated with the growth in GDP, minus efficiencies in processing.
New apps, on the other hand, are hungry, but they are much more distributed, more focused on mobile clients, and more amenable to what OrionX calls “High-Density Processing”: algorithms that have a high ops/bytes ratio running on hardware that provides similarly high ops/byte capability – ie, compute accelerators like GPUs, FPGAs, vector processors, manycore CPUs, ASICs, and new chips on the horizon.
On top of that, there will be more In-Situ Processing: processing the data wherever it happens to be, say locally on the client vs sending it around to the backend. This will be made easier by the significant rise in client-side computing power and more capable data center switches and storage nodes that can do a lot of local processing.
We will also continue to see cloud computing and virtualization eliminate idle servers and increase the utilization rates of existing systems.
Finally, commoditization of servers and racks, driven by fewer-but-larger buyers and standardization efforts like the Open Compute Project, put pressure on server costs and limit the areas in which server vendors can add value. The old adage in servers: “I know how to build it so it costs $1m, but don’t know how to build it so it’s worth $1m” will be true more than ever.
These will all combine to keep server revenues in check. We will see 5G’s wow-speeds but modest roll-out, and though it can drive a jump in video and some server-heavy apps, we’ll have to wait a bit longer.
The real battle in server architecture will be between Intel’s in-house coalition and what my colleague Dan Olds has called the Rebel Alliance: IBM’s OpenPower industry coalition. Intel brings its all-star team: Xeon Phi, Altera, Omni-Path (plus Nervana/Movidius) and Lustre, while OpenPower counters with a dream team if its own: POWER, Nvidia, Xilinx, and Mellanox (plus TrueNorth) and GPFS (Spectrum Scale). Then there is Watson which will become more visible as a differentiator, and a series of acquisitions by both camps as they fill various gaps.
The all-in-house model promises seamless integration and consistent design, while the extended team offers a best-of-breed approach. Both have merits. Both camps are pretty formidable. And there is real differentiation in strategy, design, and implementation. Competition is good.
One day in the not too distant future, your fancy car will break down on the highway for no apparent reason. It will turn out that the auto entertainment system launched a denial of service attack on the rest of the car, in a bold attempt to gain control. It even convinced the nav system to throw in with it! This is a joke, of course, but could become all too real if/when human and AI hackers get involved.
IoT is coming, and with it will come all sorts of security issues. Do you know where your connected devices are? Can you really trust them?
For the first time in decades, there is a real market opening for new chips. What drove this includes:
This will be a year when many new chips became available and tested, and there is a pretty long list of them, showing just how big the opportunity is, how eager investors must have been to not miss out, and how many different angles there are.
In addition to AI, there are a few important general-purpose chips being built. The coolest one is by startup Rex Computing, which is working on a chip for exascale, focused on very high compute-power/electric-power ratios. Qualcomm and Caviuum have manycore ARM processors, Oracle is advancing SPARC quite well, IBM’s POWER continues to look very interesting, and Intel and AMD push the X86 envelope nicely.
With AI chips, Intel already has acquired Nervana and Movidius. Google has its TPU, and IBM its neuromorphic chip, TrueNorth. Other AI chip efforts include Mobileye (the company is being bought by Intel since this article was written), Graphcore, BrainChip, TeraDeep, KnuEdge, Wave Computing, and Horizon Robotics. In addition, there are several well-publicized and respected projects like NeuRAM3, P-Neuro, SpiNNaker, Eyeriss, and krtkl going after different parts of the market. That’s a lot of chips, but most of these bets, of course, won’t pay off.
Speaking of chips, ARM servers will remain important but elusive. They will continue to make a lot of noise and point to significant wins and systems, but fail to move the needle when it comes to revenue market share in 2017.
As a long-term play, ARM is an important phenomenon in the server market – more so now with the backing of SoftBank, a much larger company apparently intent on investing and building, and various supercomputing and cloud projects that are great proving grounds.
But at the end, you need differentiation and ARM has the same problem against X86 as Intel’s Atom had against ARM. Atom did not differentiate vs ARM any more than ARM is differentiating vs Xeon.
Most systems end up being really good at something, however, and there are new apps and an open-source stack to support existing apps, which will help find specific workloads where ARM implementations could shine. That will help the differentiation become more than “it’s not X86”.
What is going on with big new apps? They keep getting more modular (microservices), more elastic (scale out), and more real-time (streaming). They’ve become their own large graph, in the computer science sense, and even more so with IoT (sensors everywhere plus In-Situ Processing).
As apps became services, they began resembling a network of interacting modules, depending on each other and evolving together. When the number of interacting pieces keeps increasing, you’ve got yourself a fabric. But as an app, it’s the kind of fabric that evolves and has an overall purpose (semantic web).
Among engineering disciplines, software engineering already doesn’t have a great reputation for predictability and maintainability. More modularity is not going to help with that.
But efforts to manage interdependence and application evolution have already created standards for structured modularity like the OSGi Alliance for Java. Smart organizations have had ways to reduce future problems (technical debt) from the get-go. So, it will be nice to see that type of methodology get better recognized and better generalized.
OK, we stop here now, leaving several interesting topics for later.
What do you think?
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.
What does a Spanish silver dollar have to do with Deep Learning? It’s a question of standards and required precision.The widely used Spanish coin was introduced at the end of the 16th century as Spain exploited the vast riches of New World silver. It was denominated as 8 Reales. Because of its standard characteristics it served as a global currency.
Ferdinand VI Silver peso (8 Reales, or Spanish silver dollar)
The American colonies in the 18th century suffered from a shortage of British coinage and used the Spanish dollar widely; it entered circulation through trade with the West Indies. The Spanish dollar was also known as “pieces of eight” and in fact was often cut into pieces known as “bits” with 8 bits comprising a dollar. This is where the expression “two bits” referring to a quarter dollar comes from. The original US dollar coin was essentially based on the Spanish dollar.
For Deep Learning, the question arises – what is the requisite precision for robust performance of a multilayer neural network. Most neural net applications are implemented with 32 bit floating point precision, but is this really necessary?
5 layer network, http://neuralnetworksanddeeplearning.com/chap6.html, CC3.0-BY-NC
It seems that many neural net applications could be successfully deployed with integer or fixed point arithmetic rather than floating point, and with only 8 to 16 bits of precision. Training may require higher precision, but not necessarily.
A team of researchers from IBM’s Watson Labs and Almaden Research Center find that:
“deep networks can be trained using only 16-bit wide fixed-point number representation when using stochastic rounding, and incur little to no degradation in the classification accuracy”.
In his blog entry “Why are Eight Bits Enough for Deep Neural Networks”, Pete Warden states that:
“As long as you accumulate to 32 bits when you’re doing the long dot products that are the heart of the fully-connected and convolution operations (and that take up the vast majority of the time) you don’t need float though, you can keep all your inputs and output as eight bit. I’ve even seen evidence that you can drop a bit or two below eight without too much loss! The pooling layers are fine at eight bits too, I’ve generally seen the bias addition and activation functions (other than the trivial relu) done at higher precision, but 16 bits seems fine even for those.”
He goes on to say that “training can also be done at low precision. Knowing that you’re aiming at a lower-precision deployment can make life easier too, even if you train in float, since you can do things like place limits on the ranges of the activation layers.”
Moussa and co-researchers have found 12 times greater speed using a fixed-point representation when compared to floating point on the same Xilinx FPGA hardware. If one can relax the precision of neural nets when deployed and/or during training, then higher performance may be realizable at lower cost and with a lower memory footprint and lower power consumption. The use of heterogeneous architectures employing GPUs, FPGAs or other special purpose hardware becomes even more feasible.
For example nVidia has just announced a system they call the “World’s first Deep Learning Supercomputer in a Box”. nVidia’s GIE inference engine software supports 16 bit floating point representations. Microsoft is pursuing the FPGA route, claiming that FGPA designs provide competitive performance for substantially less power.
This is such an interesting area, with manycore chips such as Intel’s Xeon Phi, nVidia’s GPUs and various FPGAs jockeying for position in the very hot Deep Learning marketplace.
OrionX will continue to monitor AI and Deep Learning developments closely.
References:
Gupta, S., Agrawal, A., Gopalakrishnan, K., Narayanan, P. 2015,https://arxiv.org/pdf/1502.02551.pdf “Deep Learning with Limited Numerical Precision”
Jin, L. et al. 2014, “Training Large Scale Deep Neural Networks on the Intel Xeon Phi Many-Core Processor”, proceedings, Parallel & Distributed Processing Symposium Workshops, May 2014
Moussa, M., Areibi, S., and Nichols, K. 2006 “Arithmetic Precision for Implementing BP Networks on FPGA: A case study”, Chapter 2 in FPGA Implementations of Neural Networks, ed. Omondi, A. and Rajapakse, J.
Warden, P. 2015 https://petewarden.com/2015/05/23/why-are-eight-bits-enough-for-deep-neural-networks/
Stephen Perrenod has lived and worked in Asia, the US, and Europe and possesses business experience across all major geographies in the Asia-Pacific region. He specializes in corporate strategy for market expansion, and cryptocurrency/blockchain on a deep foundation of high performance computing (HPC), cloud computing and big data. He is a prolific blogger and author of a book on cosmology.
Vector processing was an unexpected topic to emerge from the International Supercomputing Conference (ISC)held last week.
On the Monday of the conference, a new leader on the TOP500 list was announced. The Sunway TaihuLight system uses a new processor architecture that is Single-Instruction-Multiple-Data (SIMD) with a pipeline that can do eight 64-bit floating-point calculations per cycle.
This started us thinking about vector processing, a time-honored system architecture that started the supercomputing market. When microprocessors advanced enough to enable massively parallel processing (MPP) systems and then Beowulf and scale-out clusters, the supercomputing industry moved away from vector processing and led the scale-out model.
Later that day, at the “ISC 2016 Vendor Showdown”, NEC had a presentation about its project “Aurora”. This project aims to combine x86 clusters and NEC’s vector processors in the same high bandwidth system. NEC has a long history of advanced vector processors with its SX architecture. Among many achievements, it built the Earth Simulator, a vector-parallel system that was #1 on the TOP500 list from 2002 to 2004. At its debut, it had a substantial (nearly 5x) lead over the previous #1.
Close integration of accelerator technologies with the main CPU is, of course, a very desirable objective. It improves programmability and efficiency. Along those lines, we should also mention the Convey system, which goes all the way, extending the X86 instruction set, and performing the computationally intensive tasks in an integrated FPGA.
A big advantage of vector processing is that it is part of the CPU with full access to the memory hierarchy. In addition, compilers can do a good job of producing optimized code. For many codes, such as in climate modelling, vector processing is quite the right architecture.
Vector parallel systems extended the capability of vector processing and reigned supreme for many years, for very good reasons. But MPPs pushed vector processing back, and GP-GPUs pushed it further still. GPUs leverage the high volumes that the graphics market provides and can provide numerical acceleration with some incremental engineering.
But as usual, when you scale more and more, you scale not just capability, but also complexity! Little inefficiencies start adding up until they become a serious issue. At some point, you need to revisit the system and take steps, perhaps drastic steps. The Sunway TaihuLight system atop the TOP500 list is an example of this concept. And there are new applications like deep learning that look like they could use vectors to quite significant advantage.
There are other efforts towards building new exascale-class CPUs such, the “Neo processor” that Rex Computing is developing.
Will vector processing re-emerge as a lightweight, high-performing architecture?
What is the likelihood?
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.
The world of quantum computing frequently seems as bizarre as the alternate realities created in Lewis Carroll’s masterpieces “Alice’s Adventures in Wonderland” and “Through the Looking-Glass”. Carroll (Charles Lutwidge Dodgson) was a well-respected mathematician and logician in addition to being a photographer and enigmatic author.
Has quantum computing’s time actually come or are we just chasing rabbits?
That is probably a twenty million dollar question by the time a D-Wave 2X™ System has been installed and is in use by a team of researchers. Publicly disclosed installations currently include Lockheed Martin, NASA’s Ames Research Center and Los Alamos National Laboratory.
Hosted at NASA’s Ames Research Center in California, the Quantum Artificial Intelligence Laboratory (QuAIL) supports a collaborative effort among NASA, Google and the Universities Space Research Association (USRA) to explore the potential for quantum computers to tackle optimization problems that are difficult or impossible for traditional supercomputers to handle. Researchers on NASA’s QuAIL team are using the system to investigate areas where quantum algorithms might someday dramatically improve the agency’s ability to solve difficult optimization problems in aeronautics, Earth and space sciences, and space exploration. For Google the goal is to study how quantum computing might advance machine learning. The USRA manages access for researchers from around the world to share time on the system.
Using quantum annealing to solve optimization problems
D-Wave’s quantum annealing technology addresses optimization and probabilistic sampling problems by framing them as energy minimization problems and exploiting the properties of quantum physics to identify the most likely outcomes or as a probabilistic map of the solution landscape.
Quantum annealer dynamics are dominated by paths through the mean field energy landscape that have the highest transition probabilities. Figure 1 shows a path that connects local minimum A to local minimum D.
Figure 2 shows the effect of quantum tunneling (in blue) to reduce the thermal activation energy needed to overcome the barriers between the local minima with the greatest advantage observed from A to B and B to C, and a negligible gain from C to D. The principle and benefits are explained in detail in the paper “What is the Computational Value of Finite Range Tunneling?”
The D-Wave 2X: Interstellar Overdrive – How cool is that?
As a research area, quantum computing is highly competitive, but if you want to buy a quantum computer then D-Wave Systems , founded in 1999, is the only game in town. Quantum computing is as promising as it is unproven. Quantum computing goes beyond Moore’s law since every quantum bit (qubit) doubles the computational power, similar to the famous wheat and chessboard problem. So the payoff is huge, even though it is expensive, unproven, and difficult to program.
The advantage of quantum annealing machines is they are much simpler to build than gate-model quantum computing machines. The latest D-Wave machine (the D-Wave 2X), installed at NASA Ames, is approximately twice as powerful (in a quantum, exponential sense) as the previous model at over 1,000 qubits (1,097). This compares with roughly 10 qubits for current gate-model quantum systems, so two orders of magnitude. It’s a question of scale, no simple task, and a unique achievement. Although quantum researchers initially questioned whether the D-Wave system even qualified as a quantum computer, albeit a subset of quantum computing architectures, that argument seems mostly settled and it is now generally accepted that quantum characteristics have been adequately demonstrated.
In a D-Wave system, a coupled pair of qubits (quantum bits) demonstrate quantum entanglement (they influence each other), so that the entangled pair can be in any one of four states resulting from how the coupling and energy biases are programmed. By representing the problem to be addressed as an energy map the most likely outcomes can be derived by identifying the lowest energy states.
A lattice of approximately 1,000 tiny superconducting circuits (the qubits) is chilled close to absolute zero to deliver quantum effects. A user models a problem into a search for the lowest point in a vast energy landscape. The processor considers all possibilities simultaneously to determine the lowest energy required to form those relationships. Multiple solutions are returned to the user, scaled to show optimal answers, in an execution time of around 20 microseconds, practically instantaneously for all intents and purposes.
The D-wave system cabinet – “The Fridge”– is a closed cycle dilution refrigerator. The superconducting processor itself generates no heat, but to operate reliably must be cooled to about 180 times colder than interstellar space, approximately 0.015° Kelvin.
Environmental considerations: Green is the color
To function reliably, quantum computing systems require environments that are not only shielded from the Earth’s natural environment, but would be considered inhospitable to any known life form. A high vacuum is required, a pressure 10 billion times lower than atmospheric pressure, and shielded to 50,000 times less than Earth’s magnetic field. Not exactly a normal office, datacenter, or HPC facility environment.
On the other hand, the self-contained “Fridge” and servers consume just 25kW of power (approximately the power draw of a single heavily populated standard rack) and about three orders of magnitude (1000 times) less power than the current number one system on the TOP500, including its cooling system. Perhaps a more significant consideration is that power demand is not anticipated to increase significantly as it scales to several thousands of qubits and beyond.
In addition to doubling the number of qubits compared with the prior D-Wave system, the D-Wave 2X delivers lower noise in qubits and couples, delivering greater confidence in achieved results.
So much for the pictures, what about the conversations?
Now that we have largely moved beyond the debate of whether a D-Wave system is actually a quantum machine or not, then the question “OK, so what now?” could bring us back to chasing rabbits, although this time inspired by the classic Jefferson Airplane song, “White Rabbit”:
“One algorithm makes you larger
And another makes you small
But do the ones a D-Wave processes
Do anything at all?”
That of course, is where the conversations begin. It may depend upon the definition of “useful” and also a comparison between “conventional” systems and quantum computing approaches. Even the fastest supercomputer we can build using the most advanced traditional technologies can still only perform analysis by examining each possible solution serially, one solution at a time. This makes optimizing complex problems with a large number of variables and large data sets a very time consuming business. By comparison, once a problem has been suitably constructed for a quantum computer it can explore all the possible solutions at once and instantly identify the most likely outcomes.
If we consider relative performance then we begin to have a simplistic basis for comparison, at least for execution times. The QuAIL system was benchmarked for the time required to find the optimal solution with 99% probability for different problem sizes up to 945 variables. Simulated Annealing (SA), Quantum Monte Carlo (QMC) and the D-Wave 2X were compared. Full details are available in the paper referenced previously. Shown in the chart are the 50th, 75th and 85th percentiles over a set of 100 instances. The error bars represent 95% confidence intervals from bootstrapping.
This experiment occupied millions of processor cores for several days to tune and run the classical algorithms for these benchmarks. The runtimes for the higher quantiles for the larger problem sizes for QMC were not computed because the computational cost was too high.
The results demonstrate a performance advantage to the quantum annealing approach by a factor of 100 million compared with simulated annealing running on a single state of the art processor core. By comparison the current leading system on the TOP500 has fewer than 6 million cores of any kind, implying a clear performance advantage for quantum annealing based on execution time.
The challenge and the next step is to explore the mapping of real world problems to quantum machines and to improve the programming environments, which will no doubt take a significant amount of work and many conversations. New players will become more visible, early use cases and gaps will become better defined, new use cases will be identified, and a short stack will emerge to ease programming. This is reminiscent of the early days of computing or space flight.
A quantum of solace for the TOP500: Size still matters.
Even though we don’t expect to see viable exascale systems this decade, and quite likely not before the middle of the next, we won’t be seeing a Quantum500 anytime soon either. NASA talks about putting humans on Mars sometime in the 2030s and it isn’t unrealistic to think about practical quantum computing as being on a similar trajectory. Recent research at the University of New South Wales (UNSW) in Sidney, Australia demonstrated that it may be possible to create quantum computer chips that could store thousands, even millions of qubits on a single silicon processor chip leveraging conventional computer technology.
Although the current D-Wave 2X system is a singular achievement it is still regarded as being relatively small to handle real world problems, and would benefit from getting beyond pairwise connectivity, but that isn’t really the point. It plays a significant role in research into areas such as vision systems, artificial intelligence and machine learning alongside its optimization capabilities.
In the near term, we’ve got enough information and evidence to get the picture. It will be the conversations that become paramount with both conventional and quantum computing systems working together to develop better algorithms and expand the boundaries of knowledge and achievement.
References and attributions:
Paper on What is the Computational Value of Finite Range Tunneling? http://arxiv.org/pdf/1512.02206v4.pdf
Paper on Benchmarking a quantum annealing processor with the time-to-target metric: http://www.dwavesys.com/sites/default/files/ttt_arxiv.pdf
White rabbit from Sir John Tenniel’s illustrations of “Alice’s Adventures in Wonderland” by Lewis Carroll
“White Rabbit” Lyrics copyright Jefferson Airplane/ Grace Slick
Peter ffoulkes – Partner at OrionX.net
Peter is an expert in Cloud Computing, Big Data, and HPC markets, technology trends, and customer requirements which he blends to craft growth strategies and assess opportunities.
The Internet of Things (IoT) marketplace is expected to run into the trillions of dollars during the next 5 years. Gartner estimates 6.4 billion devices will be part of the Internet of Things this year. And they project $3 trillion of endpoint spending by the year 2020. [updated link] Gartner estimates 5.8 billion IoT endpoints will be in use in 2020 with revenue from endpoint electronics totaling $389 billion globally.
In our 2016 predictions blog, we said “If you missed the boat on cloud, you can’t miss it on IoT too”, and “IoT is where Big Data Analytics, Cognitive Computing, and Machine Learning come together for entirely new ways of managing business processes.”
IoT represents the ultimate convergence theme in the marketplace today. IoT includes fixed and mobile edge devices (fog computing), cloud computing, big data, analytics and machine learning, and in many instances can include social and mobile computing as well.
Because of the very large number of devices being incorporated in IoT solutions and because of the high data rates that must be supported at the very edge of a network, IoT also requires embedded computing with low-power processors for preliminary data ingestion and filtering. Data processing at the edge accomplishes two things: firstly, it allows elimination of data that does not need to be transmitted to the central cloud-based data lake. Secondly, it supports preliminary real-time processing of acquired data that can be used for device monitoring and control with immediate feedback and very low latency. Additionally, computing at the edge should be a prerequisite for enabling security of devices and the data at the time of acquisition.
IOT and cloud computing fit together like a hand in a glove. This is only natural with the highly distributed and very dynamic nature of IoT devices and IoT data. Cloud infrastructure provides the most versatile, flexible and adaptable platform for an IoT repository and processing system. Cloud vendors including Amazon (AWS), Microsoft (Azure), IBM (SoftLayer) and Alphabet (Google Compute Engine) are racing to implement IoT capabilities, APIs, and advanced analytics and machine learning solutions in their public cloud environments. For reasons of control and to meet privacy and regulatory requirements, many companies will choose to implement private cloud services as well. Hybrid cloud services will play an important role in IoT.
When it comes to infrastructure within IoT clouds and at the edge, one can imagine every type of processor, network and protocol, servers and storage playing a role – both in distributed and centralized fashions – for IoT environments. Shared APIs and protocols are critical to promote interoperability. End-to-end security is as well an imperative; security is needed within every device and at each and every layer and sub-layer of the solution.
If anything deserves the term Big Data it is IoT. Billions and billions of devices are becoming Internet-enabled and the number of devices engaged in IoT applications will soon exceed the number of people on Earth. Because there are now many Big Data solutions available in public clouds this reinforces the cloud as a natural repository for IoT-generated Big Data lakes.
Analytics and machine learning will be heavily used to extract maximum value from the large amount of data acquired. A very wide range of business intelligence, analytics and machine learning techniques will be required. We foresee that IoT will be a big driver for machine learning advances. Analytics and machine learning will support everything from better operations of the connected devices to better information on how devices are utilized and new value-added services enabled by the device manufacturers.
There is a wide array of potential IoT applications covering every industry: Transportation, Retail, Manufacturing, Energy, Finance, Smart Cities, Healthcare, Agriculture, Government and other areas. Consumer apps are found in areas such as wearables, fitness, smart homes, electronics, and gaming.
Benefits from IoT applications will include improved operational efficiency and reliability, insight into usage patterns, and opportunities for increased revenue and better product design for manufacturers. Just as today we cannot imagine life without the Internet, we will not be able to imagine life without IoT by the beginning of the next decade.
It’s difficult to think of an area that requires a more holistic and converged view than IoT and that is not already a natural subset or example of IoT. Robotics? Drones? 3D printing? …. Every device of value should benefit from being plugged into the Internet, if only for maintenance and monitoring purposes, and these IoT-enabled devices will become available to participate in other IoT apps. At the other end of the data flow, big value comes from analytics services that are applied to a pool of devices and the streaming data they provide.
IT and IoT will become inseparable. IoT requires, incorporates, and benefits from all of the other major themes in IT today and thus we see it as representing the Ultimate Convergence at present.
What is your IoT strategy? OrionX can help.
Stephen Perrenod has lived and worked in Asia, the US, and Europe and possesses business experience across all major geographies in the Asia-Pacific region. He specializes in corporate strategy for market expansion, and cryptocurrency/blockchain on a deep foundation of high performance computing (HPC), cloud computing and big data. He is a prolific blogger and author of a book on cosmology.
Here at OrionX.net, we are fortunate to work with tech leaders across several industries and geographies, serving markets in Mobile, Social, Cloud, and Big Data (including Analytics, Cognitive Computing, IoT, Machine Learning, Semantic Web, etc.), and focused on pretty much every part of the “stack”, from chips to apps and everything in between. Doing this for several years has given us a privileged perspective. We spent some time to discuss what we are seeing and to capture some of the trends in this blog: our 2016 technology issues and predictions. We cut it at 17 but we hope it’s a quick read that you find worthwhile. Let us know if you’d like to discuss any of the items or the companies that are driving them.
Energy is arguably the most important industry on the planet. Advances in energy efficiency and sustainable energy sources, combined with the debate and observations of climate change, and new ways of managing capacity risk are coming together to have a profound impact on the social and political structure of the world, as indicated by the Paris Agreement and the recent collapse in energy prices. These trends will deepen into 2016.
Money was the original virtualization technology! It decoupled value from goods, simplified commerce, and enabled the service economy. Free from the limitations of physical money, cryptocurrencies can take a fresh approach to simplifying how value (and ultimately trust, in a financial sense) is represented, modified, transferred, and guaranteed in a self-regulated manner. While none of the existing implementations accomplish that, they are getting better understood and the ecosystem built around them will point the way toward a true digital currency.
Whether flying, driving, walking, sailing, or swimming, drones and robots of all kinds are increasingly common. Full autonomy will remain a fantasy except for very well defined and constrained use cases. Commercial success favors technologies that aim to augment a human operator. The technology complexity is not in getting one of them to do an acceptable job, but in managing fleets of them as a single network. But everything will be subordinate to an evolving and complex legal framework.
A whole new approach to computing (as in, not binary any more), quantum computing is as promising as it is unproven. Quantum computing goes beyond Moore’s law since every quantum bit (qubit) doubles the computational power, similar to the famous wheat and chessboard problem. So the payoff is huge, even though it is, for now, expensive, unproven, and difficult to use. But new players will become more visible, early use cases and gaps will become better defined, new use cases will be identified, and a short stack will emerge to ease programming. This is reminiscent of the early days of computing so a visit to the Computer History Museum would be a good recalibrating experience.
The changing nature of work and traditional jobs received substantial coverage in 2015. The prospect of artificial intelligence that could actually work is causing fears of wholesale elimination of jobs and management layers. On the other hand, employers routinely have difficulty finding talent, and employees routinely have difficulty staying engaged. There is a structural problem here. The “sharing economy” is one approach, albeit legally challenged in the short term. But the freelance and outsourcing approach is alive and well and thriving. In this model, everything is either an activity-sliced project, or a time-sliced process, to be performed by the most suitable internal or external resources. Already, in Silicon Valley, it is common to see people carrying 2-3 business cards as they match their skills and passions to their work and livelihood in a more flexible way than the elusive “permanent” full-time job.
With the tectonic shifts in technology, demographic, and globalization, companies must transform or else. Design thinking is a good way to bring customer-centricity further into a company and ignite employees’ creativity, going beyond traditional “data driven needs analysis.” What is different this time is the intimate integration of arts and sciences. What remains the same is the sheer difficulty of translating complex user needs to products that are simple but not simplistic, and beautiful yet functional.
Old guard IT vendors will have the upper hand over new Cloud leaders as they all rush to claim IoT leadership. IoT is where Big Data Analytics, Cognitive Computing, and Machine Learning come together for entirely new ways of managing business processes. In its current (emerging) condition, IoT requires a lot more vertical specialization, professional services, and “solution-selling” than cloud computing did when it was in its relative infancy. This gives traditional big (and even small) IT vendors a chance to drive and define the terms of competition, possibly controlling the choice of cloud/software-stack.
Cybercrime is big business and any organization with digital assets is vulnerable to attack. As Cloud and Mobile weaken IT’s control and IoT adds many more points of vulnerability, new strategies are needed. Cloud-native security technologies will include those that redirect traffic through SaaS-based filters, Micro-Zones to permeate security throughout an app, and brand new approaches to data security.
In any value chain, a vendor must decide what value it offers and to whom. With cloud computing, the IT value chain has been disrupted. What used to be a system is now a piece of some cloud somewhere. As the real growth moves to “as-a-service” procurements, there will be fewer but bigger buyers of raw technology who drive hardware design towards scale and commoditization.
The computing industry was down the path of hardware partitioning when virtualization took over, and dynamic reconfiguration of hardware resources took a backseat to manipulating software containers. Infrastructure-as-code, composable infrastructure, converged infrastructure, and rack-optimized designs expand that concept. But container reconfiguration is insufficient at scale, and what is needed is hardware reconfiguration across the data center. That is the next frontier and the technologies to enable it are coming.
Mobile devices are now sufficiently capable that new features may or may not be needed by all users and new OS revs often slow down the device. Even with free upgrades and pre-downloaded OS revs, it is hard to make customers upgrade, while power users jailbreak and get the new features on an old OS. Over time, new capabilities will be provided via more modular dynamically loaded OS services, essentially a new class of apps that are deeply integrated into the OS, to be downloaded on demand.
Nowhere are the demands for scale, flexibility and effectiveness for analytics greater than in social media. This is far beyond Web Analytics. The seven largest “populations” in the world are Google, China, India, Facebook, WhatsApp, WeChat and Baidu, in approximately that order, not to mention Amazon, Apple, Samsung, and several others, plus many important commercial and government applications that rely on social media datasets. Searching through such large datasets with very large numbers of images, social commentary, and complex network relationships stresses the analytical algorithms far beyond anything ever seen before. The pace of algorithmic development for data analytics and for machine intelligence will accelerate, increasingly shaped by social media requirements.
Legacy modernization will get more attention as micro-services, data-flow, and scale-out elasticity become established. But long-term, software engineering is in dire need of the predictability and maintainability that is associated with other engineering disciplines. That need is not going away and may very well require a wholesale new methodology for programming. In the meantime, technologies that help automate software modernization, or enable modular maintainability, will gain traction.
The technical and operational burden on developers has been growing. It is not sustainable. NoSQL databases removed the time-delay and complexity of a data schema at the expense of more complex codes, pulling developers closer to data management and persistence issues. DevOps, on the other hand, has pulled developers closer to the actual deployment and operation of apps with the associated networking, resource allocation, and quality-of-service (QoS) issues. This is another “rubber band” that cannot stretch much more. As cloud adoption continues, development, deployment, and operations will become more synchronized enabling more automation.
The idea of a “memory-only architecture” dates back several decades. New super-large memory systems are finally making it possible to hold entire data sets in memory. Combine this with Flash (and other emerging storage-class memory technologies) and you have the recipe for entirely new ways of achieving near-real-time/straight-through processing.
Small and mid-size public cloud providers will form coalitions around a large market leader to offer enterprise customers the flexibility of cloud without the lock-in and the risk of having a single supplier for a given app. This opens the door for transparently running a single app across multiple public clouds at the same time.
It’s been years since most app developers needed to know what CPU their app runs on, since they work on the higher levels of a tall software stack. Porting code sill requires time and effort but for elastic/stateless cloud apps, the work is to make sure the software stack is there and works as expected. But the emergence of large cloud providers is changing the dynamics. They have the wherewithal to port any system software to any CPU thus assuring a rich software infrastructure. And they need to differentiate and cut costs. We are already seeing GPUs in cloud offerings and FPGAs embedded in CPUs. We will also see the first examples of special cloud regions based on one or more of ARM, OpenPower, MIPS, and SPARC. Large providers can now offer a usable cloud infrastructure using any hardware that is differentiated and economically viable, free from the requirement of binary compatibility.
OrionX is a team of industry analysts, marketing executives, and demand generation experts. With a stellar reputation in Silicon Valley, OrionX is known for its trusted counsel, command of market forces, technical depth, and original content.
There has been much discussion in recent years as to the continued relevance of the High Performance Linpack (HPL) benchmark as a valid measure of the performance of the world’s most capable machines, with some supercomputing sites opting out of the TOP500 completely such as National Center for Supercomputing Applications at the University of Illinois at Urbana-Champaign that houses the Cray Blue Waters machine. Blue Waters is claimed to be the fastest supercomputer on a university campus, although without an agreed benchmark it is a little hard to verify such a claim easily.
At a superficial level it appears as though any progress in the upper echelons of the TOP500 may have stalled with little significant change in the TOP10 in recent years and just two new machines added to that elite group in the November 2015 list: Trinity at 8.10 Pflop/s and Hazel Hen at 5.64 Pflop/s, both Cray systems increasing the company’s share of the TOP10 machines to 50%. The TOP500 results are available at http://www.top500.org.
A new benchmark – the High Performance Conjugate Gradients (HPCG) – introduced two years ago was designed to be better aligned with modern supercomputing workloads and includes system features such as high performance interconnects, memory systems, and fine grain cooperative threading. The combined pair of results provided by HPCG and HPL can act as bookends on the performance spectrum of any given system. As of July 2015 HPCG had been used to rank about 40 of the top systems on the TOP500 list.
The November TOP500 List: Important Data for Big Iron
Although being complemented by other benchmarks, the HPL-based TOP500 is still the most significant resource for tracking the pace of super computing technology development historically and for the foreseeable future.
The first petascale machine, ‘Roadrunner’, debuted in June 2008, twelve years after the first terascale machine – ASCI Red in 1996. Until just a few years ago 2018 was the target for hitting the exascale capability level. As 2015 comes to its close the first exascale machine seems much more likely to debut in the first half of the next decade and probably later in that window rather than earlier. So where are we with the TOP500 this November, and what can we expect in the next few lists?
Highlights from the November 2015 TOP500 List on performance:
– The total combined performance of all 500 systems had grown to 420 Pflop/s, compared to 361 Pflop/s last July and 309 Pflop/s a year previously, continuing a noticeable slowdown in growth compared to the previous long-term trend.
– Just over 20% of the systems (a total of 104) use accelerator/co-processor technology, up from 90 in July 2015
– The latest list shows 80 petascale systems worldwide making a total of 16% of the TOP500, up from 50 a year earlier, and significantly more than double the number two years earlier. If this trend continues the entire TOP100 systems are likely to be petascale machines by the end of 2016.
So what do we conclude from this? Certainly that the road to exascale is significantly harder than we may have thought, not just from a technology perspective, but even more importantly from a geo-political and commercial perspective. The aggregate performance level of all of the TOP500 machines has just edged slightly over 40% of the HPL metric for an exascale machine.
Most importantly, increasing the number of petascale-capable resources available to scientists, researchers, and other users up to 20% of the entire list will be a significant milestone. From a useful outcome and transformational perspective it is much more important to support advances in science, research and analysis than to ring the bell with the world’s first exascale system on the TOP500 in 2018, 2023 or 2025.
Architecture
As acknowledged by the introduction of HPCG, HPL and the TOP500 performance benchmark are only one part of the HPC performance equation.
The combination of modern multi-core 64 bit CPUs and math accelerators from Nvidia, Intel and others have addressed many of the issues related to computational performance. The focus on bottlenecks has shifted away from computational strength to data-centric and energy issues, which from a performance perspective influence HPL results but are not explicitly measured in the results. However, from an architectural perspective the TOP500 lists still provide insight into the trends in an extremely useful manner.
Observations from the June 2015 TOP500 List on system interconnects:
– After two years of strong growth from 2008 to 2010, InfiniBand-based TOP500 systems plateaued at 40% of the TOP500 while compute performance grew aggressively with the focus on hybrid, accelerated systems.
– The uptick in InfiniBand deployments June 2015 to over 50% of the TOP500 list for the first time does not appear to be the start of a trend, with Gigabit Ethernet systems increasing to over a third (36.4%) of the TOP500, up from 29.4%.
The TOP500: The November List and What’s Coming Soon
Although the TOP10 on the list have shown little change for a few years, especially with the planned upgrade to Tianhe 2 blocked by the US government, there are systems under development in China that are expected to challenge the 100 Pflop/s barrier in the next twelve months. From the USA, Coral initiative is expected to significantly exceed the 100 Pflop/s limit – targeting the 100 to 300 Pflop/s with two machines from IBM and one from Cray, but not before the 2017 time frame. None of these systems are expected to deliver more than one third of exascale capability.
Vendors: A Changing of the Guard
The USA is still the largest consumer of TOP500 systems with 40% of the share by system count, which clearly favors US vendors, but there are definite signs of change on the wind.
Cray is the clear leader in performance with a 24.9 percent share of installed total performance, roughly the same as a year earlier. Each of the other leading US vendors is experiencing declining share with IBM in second place with a 14.9 percent share, down from 26.1 percent last November. Hewlett Packard is third with 12.9 percent, down from 14.5 percent twelve months ago.
These trends are echoed in the system count numbers. HP has the lead in systems and now has 156 (31%), down from 179 systems (36%) twelve months ago. Cray is in second place with 69 systems (13.8%), up from 12.4% last November. IBM now has 45 systems (9%), down from 136 systems (27.2%) a year ago.
On the processor architecture front a total of 445 systems (89%) are now using Intel processors, slightly up from 86.2 percent six months ago, with IBM Power processors now at 26 systems (5.2%), down from 38 systems (7.6%) six month ago.
Geo-Political Considerations
Perhaps the single most interesting insight to come out of the TOP500 has little to do with the benchmark itself, and the geographical distribution data which in itself underscores the importance of the TOP500 lists in tracking trends.
The number of systems in the United States has fallen to the lowest point since the TOP500 list was created in 1993, down to 201 from 231 in July. The European share has fallen to 107 systems compared to 141 on the last list and is now lower than the Asian share, which has risen to 173 systems, up from 107 the previous list and with China nearly tripling the number of systems to 109. In Europe, Germany is the clear leader with 32 systems, followed by France and the UK at 18 systems each. Although the numbers for performance-based distributions vary slightly, the percentage distribution is remarkably similar.
The USA and China divide over 60% of the TOP500 systems between them, with China having half the number and capability of the US-based machines. While there is some discussion about how efficiently the Chinese systems are utilized in comparison to the larger US-based systems, that situation is likely to be temporary and could be addressed by a change in funding models. It is abundantly clear that China is much more serious about increasing funding for HPC systems while US funding levels continue to demonstrate a lack of commitment to maintaining development beyond specific areas such as high energy physics.
TOP500: “I Have Seen The Future and it Isn’t Exascale”
The nature of HPC architectures, workloads and markets are all in flux in the next few years, and an established metric such as the TOP500 is an invaluable tool to monitor the trends. Even though HPL is no longer the only relevant benchmark, the emergence of complementary benchmarks such as HPCG serves to enhance the value of the TOP500, not to diminish it. It will take a while to generate a body of data equivalent to the extensive resource that has been established by the TOP500.
What we can see from the current trends shown in the November 2015 TOP500 list is that although an exascale machine will be built eventually, it almost certainly will not be delivered this decade and possibly not until half way through the next decade. Looking at the current trends it is also very unlikely that the first exascale machine will be built in the USA, and much more likely that it will be built in China.
Over the next few years the TOP500 is likely to remain the best tool we have to monitor geo-political and technology trends in the HPC community. If the adage coined by the Council on Competitiveness, “To out-compute is to out-compete” has any relevance, the TOP500 will likely provide us with a good idea of when the balance of total compute capability will shift from the USA to China.
Peter ffoulkes – Partner at OrionX.net
Peter is an expert in Cloud Computing, Big Data, and HPC markets, technology trends, and customer requirements which he blends to craft growth strategies and assess opportunities.
Last week, in Part 1 of this two-part blog, we looked at trends in Big Data and analytics, and started to touch on the relationship with HPC (High Performance Computing). In this week’s blog we take a look at the usage of Big Data in HPC and what commercial and HPC Big Data environments have in common, as well as their differences.
High Performance Computing has been the breeding ground for many important mainstream computing and IT developments, including:
Big Data has indeed been a reality in many HPC disciplines for decades, including:
Horseshoe Falls (portion of Niagara Falls on Canadian side)
All of these fields and others generate massive amounts of data, which must be cleaned, calibrated, reduced and analyzed in great depth in order to extract knowledge. This might be a new genetic sequence, the identification of a new particle such as the Higgs Boson, the location and characteristics of an oil reservoir, or a more accurate weather forecast. And naturally the data volumes and velocity are growing continually as scientific and engineering instrumentation becomes more advanced.
A recent article, published in the July 2015 issue of the Communications of the ACM, is titled “Exascale computing and Big Data”. Authors Daniel A. Reed and Jack Dongarra note that “scientific discovery and engineering innovation requires unifying traditionally separated high-performance computing and big data analytics”.
(n.b. Exascale is 1000 x Petascale, which in turn is 1000 x Terascale. HPC and Big Data are already well into the Petascale era. Europe, Japan, China and the U.S. have all announced Exascale HPC initiatives spanning the next several years.)
What’s in common between Big Data environments and HPC environments? Both are characterized by racks and racks of commodity x86 systems configured as compute clusters. Both environments have compute system management challenges in terms of power, cooling, reliability and administration, scaling to as many as hundreds of thousands of cores and many Petabytes of data. Both are characterized by large amounts of local node storage, increasing use of flash memory for fast data access and high-bandwidth switches between compute nodes. And both are characterized by use of Linux OS operating systems or flavors of Unix. Open source software is generally favored up through the middleware level.
What’s different? Big Data and analytics uses VMs above the OS, SANs as well as local storage, the Hadoop (parallel) file system, key-value store methods, and a different middleware environment including Map-Reduce, Hive and the like. Higher-level languages (R, Python, Pig Latin) are preferred for development purposes.
HPC uses C, C++, and Fortran traditional compiler development environments, numerical and parallel libraries, batch schedulers and the Lustre parallel file system. And in some cases HPC systems employ accelerator chips such as Nvidia GPUs or Intel Xeon Phi processors, to enhance floating point performance. (Prediction: we’ll start seeing more and more of these used in Big Data analytics as well – http://www.nvidia.com/object/data-science-analytics-database.html).
But in both cases the pipeline is essentially:
Data acquisition -> Data processing -> Model / Simulation -> Analytics -> Results
The analytics must be based on and informed by a model that is attempting to capture the essence of the phenomena being measured and analyzed. There is always a model — it may be simple or complex; it may be implicit or explicit.
Human behavior is highly complex, and every user, every customer, every patient, is unique. As applications become more complex in search of higher accuracy and greater insight, and as compute clusters and data management capabilities become more powerful, the models or assumptions behind the analytics will in turn become more complex and more capable. This will result in more predictive and prescriptive power.
Our general conclusion is that while there are some distinct differences between Big Data and HPC, there are significant commonalities. Big Data is more the province of social sciences and HPC more the province of physical sciences and engineering, but they overlap, and especially so when it comes to the life sciences. Is bioinformatics HPC or Big Data? Yes, both. How about the analysis of clinical trials for new pharmaceuticals? Arguably, both again.
So cross-fertilization and areas of convergence will continue, while each of Big Data and HPC continue to develop new methods appropriate to their specific disciplines. And many of these new methods will crossover to the other area when appropriate.
The National Science Foundation believes in the convergence of Big Data and HPC and is putting $2.4 million of their money into this at the University of Michigan, in support of various applications including climate science, cardiovascular disease and dark matter and dark energy. See:
What are your thoughts on the convergence (or not) of Big Data and HPC?
Stephen Perrenod has lived and worked in Asia, the US, and Europe and possesses business experience across all major geographies in the Asia-Pacific region. He specializes in corporate strategy for market expansion, and cryptocurrency/blockchain on a deep foundation of high performance computing (HPC), cloud computing and big data. He is a prolific blogger and author of a book on cosmology.
Data volumes, velocity, and variety are increasing as consumer devices become more powerful. PCs, smart phones and tablets are the instrumentation, along with the business applications that continually capture user input, usage patterns and transactions. As devices become more powerful each year (each few months!) the generated volumes of data and the speed of data flow both increase concomitantly. And the variety of available applications and usage models for consumer devices is rapidly increasing as well.
Holding more and more data in-memory, via in-memory databases and in-memory computing, is becoming increasingly important in Big Data and data management more broadly. HPC has always required very large memories due to both large data volumes and the complexity of the simulation models.
Igauzu Falls: By Mario Roberto Duran Ortiz Mariordo (Own work) CC BY 3.0, via Wikimedia Commons
As is often pointed out in the Big Data field, it is the analytics that matters. Collecting, classifying and sorting data is a necessary prerequisite. But until a proper analysis is done, one has only expended time, energy and money. Analytics is where the value extraction happens, and that must justify the collection effort.
Applications for Big Data include customer retention, fraud detection, cross-selling, direct marketing, portfolio management, risk management, underwriting, decision support, and algorithmic trading. Industries deploying Big Data applications include telecommunications, retail, finance, insurance, health care, and the pharmaceutical industry.
There are a wide variety of statistical methods and techniques employed in the analytical phase. These can include higher-level AI or machine learning techniques e.g. neural networks, support vector machines, radial basis functions, and nearest neighbor methods. These imply a significant requirement for a large number of floating point operations, which is characteristic of most of HPC.
For one view on this, here is a recent report on InsideHPC.com and video on “Why HPC is so important to AI”
If one has the right back-end applications and systems then it is possible to keep up with the growth in data and perform the deep analytics necessary to extract new insights about customers, their wants and desires, and their behavior and buying patterns. These back-end systems increasingly need to be of the scale of HPC systems in order to stay on top of all of the ever more rapidly incoming data, and to meet the requirement to extract maximum value.
In Part 2 of this blog series, we’ll look at how Big Data and HPC environments differ, and at what they have in common.
Stephen Perrenod has lived and worked in Asia, the US, and Europe and possesses business experience across all major geographies in the Asia-Pacific region. He specializes in corporate strategy for market expansion, and cryptocurrency/blockchain on a deep foundation of high performance computing (HPC), cloud computing and big data. He is a prolific blogger and author of a book on cosmology.
The IT industry is arguably going through its biggest transformation since the introduction of the IBM PC in the early 1980s, and a cloud service based approach is the likely destination for most enterprise workloads. On August 26th and 27th Mirantis hosted the “second annual” OpenStack Silicon Valley event at the Computer History Museum in Mountain View to a standing room only audience.
Setting the scene for the conference, Alex Freedland, Co-Founder and President of Mirantis, pointed out the market momentum gained by OpenStack in five short years, displacing other potential cloud platforms and propelling it very firmly to being one of just five viable cloud computing platforms for enterprise deployment: Amazon Web Services, Google Cloud Platform, Microsoft Azure, VMware, and the OpenStack Community and Partner Ecosystem.
On the surface this might seem to be a rather bold claim for such a young open source technology offering as OpenStack, but based on my end-user studies at 451 Research prior to joining OrionX, I find the claim entirely credible. It is important to remember that despite all of the hype surrounding cloud computing, most enterprises are still in the early stages of cloud adoption and that it is a complex transition. Although the technology platforms to build a cloud are critical to success, there are many more significant factors that will contribute to success. First and foremost, the shift to a cloud-based approach to delivering IT services is a huge cultural transition and with such transitions also come substantial business risk. This was one of the key findings of my own research, and also the main theme of the talk by James Staten, Chief Strategist of the Cloud and Enterprise division at Microsoft and until recently Vice President & Principal Analyst for cloud computing at Forrester Research.
Another factor in a successful transition is choosing the right partners for technology and business solutions. It is here that OpenStack gains enormous credibility, not so much from the current level of maturity of the OpenStack technology platform but from the massive amount of investment in both money and developers that is being poured into the effort to make OpenStack successful, something that simply cannot be matched by any single company on the planet today. There is still much to be done to turn OpenStack into a true enterprise-grade, production-ready cloud platform but it is also clear that OpenStack is funded for success, which is one very important consideration for any enterprise making a significant bet on the foundation for its future IT services.
There are too many proof points to mention that underpin the level of activity and initiatives supporting the momentum behind OpenStack, but a few recent and noteworthy ones are:
So, are we all off to see the wizard, the wonderful wizard of OS [OpenStack]?
Possibly, but not quite so fast! There are still lions and tigers and bears, Oh my! There is still much work to do to bring OpenStack, and indeed all the other credible cloud platforms up to production enterprise deployment standards, and this will take time even if the money and the will is there, which it appears to be for OpenStack. Here are just a few of the initiatives outlined at the OpenStack Silicon Valley summit that point to what is in the pipeline to make OpenStack more enterprise ready:
Core Improvements:
Strategic Investments
The long and winding road
Perhaps the biggest issue of all to be addressed is the current scarcity of OpenStack expertise and developers. They seem to be worth more than their weight in gold or Bitcoins, and they also seem to know it! These are some of the driving forces behind initiatives such as Intel’s Cloud for All, and why a company like Mirantis is gaining recognition for its expertise, and investment that enhances its ability to influence the future of cloud computing.
The journey to “the cloud” is a long and twisty one, and there will be more than one destination, but for now OpenStack seems likely to be one of the five most viable destinations. From here it appears to be more real than it does to be a mirage.
Peter ffoulkes, Partner OrionX
Peter is an expert in Cloud Computing, Big Data, and HPC markets, technology trends, and customer requirements which he blends to craft growth strategies and assess opportunities.
Long-term planning is an art. But it might take a real scientist to tackle a hundred-year planning horizon!
Frank Wilczek is a Nobel Prize winner in theoretical physics, awarded for his work in quantum chromodynamics (quarks, to you and me). He is currently the Herman Feshbach Professor of Physics at M.I.T.
Professor Wilczek was invited to speak at Brown University’s 250th anniversary last year, and was asked to make predictions about the future of physics and technology 250 years from now. Considering that a much too difficult assignment, he modified (re-normalized in physics terms) the assignment to looking forward into the next century.
We won’t look here at his predictions for advancement in physics, many of which have to do with further unification of the laws of physics, such as supersymmetry, but instead focus on long-term planning for technology and his predictions in that area. His paper “Physics in 100 Years” is available here: http://arxiv.org/pdf/1503.07735.pdf and includes both the physics and technology predictions and other speculations about the future of humanity.
Here are some of his technology predictions (quotes from the paper are shown in italics) along with our elaboration, interpretation and reflections.
Calculation will increasingly replace experimentation in design of useful materials, catalysts, and drugs, leading to much greater efficiency and new opportunities for creativity.
Computation is key to the nanoscale revolution for developing super-strong yet highly flexible and lightweight materials that can be 3-D printed for a wide variety of applications. And rapid computation is key to developing new drugs specific to individuals’ genetic makeup (targeted gene therapies).
Calculation of many nuclear properties from fundamentals will reach < 1% accuracy, allowing much more accurate modeling of supernovae and of neutron stars. Physicists will learn to manipulate atomic nuclei dexterously, as they now manipulate atoms, enabling (for example) ultra-dense energy storage and ultra-high energy lasers.
Currently, all we are really able to do at the nuclear level is build fission reactors and fission or fusion bombs. As one example, hydrogen fusion will become economically viable during this century, liberating tremendous amounts of energy from deuterium and tritium (heavy hydrogen nuclei) extracted from water, but with lower associated risks as compared to nuclear reactors using uranium.
Capable three-dimensional, fault-tolerant, self-repairing computers will be developed. In engineering those features, we will learn lessons relevant to neurobiology.
The human brain is a 3-D construct with an extremely complex network and employing high degrees of parallelism, which is key to its processing speed and thus intelligence. Computers as systems are today packaged to some extent in 3-D, but are intrinsically based around 2-dimensional CPU chips and 2-D memory chips. These chips are connected one to the other with simple networks. In the future CPUs, together with their associated memory, will be designed with 3-D architectures, allowing for much faster speeds, much higher connectivity and very much greater memory bandwidth. Quantum computing technology based on more robust qubits, rather than bits, will be well-established, allowing for tremendous speedups for certain classes of algorithms. The image below is of the CPU chip from the first line of commercially available quantum computers. One of these DWave systems is operated jointly by Google and the NASA Advanced Supercomputing Facility in Mountain View, California.
Self-assembling, self-reproducing, and autonomously creative machines will be developed. Their design will adapt both ideas and physical modules from the biological world.
We’re talking intelligent robots, here, folks! And other autonomous, intelligent, and yes, self-reproducing machines, some very tiny (able to enter the bloodstream for medical purposes), others very large (see macroscale). Artificial organs. Asteroid mining machines. Robot armies (war without human casualties?). The possibilities are endless. Asimov’s 3 laws of robotics will be enforced.
Bootstrap engineering projects wherein machines, starting from crude raw materials, and with minimal human supervision, build other sophisticated machines – notably including titanic computers – will be underway.
Future supercomputers will self-assemble, with limited human oversight in the assembly process. Programming will have much higher levels of machine assistance, with problem statements represented at a very high level. Machinery and vehicle production will be almost entirely automated.
A substantial fraction of the Sun’s energy impinging on Earth will be captured for human use.
The deserts will bloom with advanced solar cells and superconducting transmission lines. Humankind will have settled and begun working in the inner solar system, including the Moon, Mars, one or more major asteroids, and one or more of Jupiter’s moons (e.g. Europa, Callisto).
Imagine how much has changed in the past 100 years. No one had flown on a commercial aircraft. Very few people had telephones or automobiles or radios. And consider that the pace of discovery, knowledge acquisition and technological development is today much higher, and still accelerating. A 100 years from now, average human lifetimes should exceed a century, based on revolutionary medical advances and ever safer transportation.
We’d be interested to hear from you, what are your thoughts on the state of technology a century from now? Please comment.
Stephen Perrenod has lived and worked in Asia, the US, and Europe and possesses business experience across all major geographies in the Asia-Pacific region. He specializes in corporate strategy for market expansion, and cryptocurrency/blockchain on a deep foundation of high performance computing (HPC), cloud computing and big data. He is a prolific blogger and author of a book on cosmology.
The industry is awash with articles on big data. Big data news is not confined to the technical webpages anymore. You can read about big data on Forbes and The Economist for example.
Each week the technical media reports on break through, startups, funding and customer use cases.
No matter your source for information on big data one thing they all have in common is that the amount of information an organization will deal with only ever increase, this is driving the ‘big data’ concept. Organizations are relying on increasing amounts of information from a variety of sources, text, images, audio, video, etc., to analyze, improve and execute their operations.
Big data is happening with Industries such as financial services, healthcare, retail, communications, to name a few. These industries have been collecting data for years and with the advent of the Internet of Things data is growing exponentially. The challenge is how to make sense of this data and turn it into business value – analytics and predictive analytics.
An example of big data at work is Fraud Detection and Security requirements of the Banking and Financial Services industry. These organizations need to prevent fraud and leverage analytics, machine learning, and big data technology to gain a holistic view of their customers to identify patterns buried in data and distinguish fraudulent activity from normal activity.
Many new solutions are being developed to churn through and make sense of this data. Perhaps an old solution is getting its second or third wind – big memory, really big memory and the ability to troll through with high bandwidth and latency speeds.
Why will this be important?
First, memory is getting cheaper and cheaper. Second, by storing the database in-memory one avoids the traditional IO bottleneck, despite the growth in SSD storage, and third there are some really big memory systems available, up to 64TB of shared memory in some cases and getting bigger. These tera-scale memory systems have the ability to sift through masses of data enabling OLTP and OLAP in a single system creating a Hybrid/Transaction Analytics Processing (HTAP) solutions that do not require data duplication.
In-memory computing is not new. For several decades the supercomputing industry or High Performance Computing squeezed every bit of performance they could with in-memory computing solutions. Now we are seeing in-memory HPC techniques being applied to everyday big data. HPC to the rescue of the enterprise, one more time.
Steve brings a unique blend of expertise to the technology industry. He has held senior VP positions in sales & marketing for HPC and Enterprise vendors including Sun Microsystems and Hitachi. He’s also worked with early stage technology and Internet start-ups helping to raise over $300M for clients.
Welcome to Orion and OrionX.net!
When we formed Orion Marketing more than 6 years ago, our vision was to offer a technology marketing capability that would cover the entire spectrum from the board room to customer programs. We anticipated that marketing as a function would be transformed in significant ways: more technology, more data, more social, more blending of arts and sciences, more integrated with every other function, and ultimately more critical to the organization.
This required a new model.
If marketing was to be so important, it needed first to have a deeper understanding of global trends, business strategy, services, technologies, sales, human resources, law, finance, operations, and geographies. That is a big task, but it is what it takes to begin to do great marketing.
So from the beginning, our focus has been on the new world of marketing. That means depth, breadth and quality. It means synthesis that must follow analysis to create great content. In short, it means everything that must be in place to “connect products to customers,” as our tagline said.
That was the right focus, and thanks to that and our wonderful customers, Orion Marketing grew, which enabled us to codify and hone our core marketing services. We also started a systematic program to build additional capabilities.
Today, we are ready to launch a significant expansion of our services based on a few years of work.
We have developed, delivered, and are now formally launching two additional services: Strategy, and PR services. This expansion warrants a new brand that recognizes our heritage and points to a future of innovation and customer value in the digital era.
OrionX continues the same unique model: bringing together business acumen, technical depth, insightful industry perspective, high quality work and counsel, and a natural focus on customer success.
Marketing has come a long way in the last decade. Just as companies have a well-practiced understanding of what it takes to build a great product, they now have a growing appreciation of why marketing is essential to their business and what it takes to do great marketing. They recognize that marketing is the glue at the center of the organization, and must have high quality content at its own center. Nearly 25 years after Regis McKenna’s article in Harvard Business Review, finally Marketing is Everything!
As we embark on the next phase of this journey, we strive to continue to deserve your support, advice, participation, and business.
Thank you!
Shahin is a technology analyst and an active CxO, board member, and advisor. He serves on the board of directors of Wizmo (SaaS) and Massively Parallel Technologies (code modernization) and is an advisor to CollabWorks (future of work). He is co-host of the popular RadioFreeHPC (“Car Talk” of supercomputing), OrionX Download, and Afero IoT podcasts.