Fast Forward: Super Computing With Lenovo's Brian Connors
Welcome to Fast Forward, where we have conversations about living in the future. I'm coming to you this week from the SAP Sapphire Now Conference. My guest is Brian Connors, the VP of Strategic Technology Alliances and Development for Lenovo's Data Center Group.
Data centers are one of the things that are making the world the amazing place it is today and Brian can tell us all about them. He is also going to speak to the history of the PC industry, artificial intelligence and automation. Read and watch our full conversation below.
Dan Costa: We should probably start by talking about where we are. People who are watching video can see that we're in a conference center, there are a lot of people around. SAP Sapphire Now is actually a huge show here in Orlando. Why don't you explain a little bit about what's going on?
Brian Connors: This is my third event here in Orlando. It's what I would consider the premier show of business, business leaders, business process, enterprise business. Two years ago at the keynote, you could stand in the back and watch. Today, it was like 10 people deep standing, plus it was being watched by everybody via streams. I think there was 30,000 people here today. For us as a data center group, we're a strategic partner of SAP and a platinum partner here at the show.
Just a little bit of history here, IBM x86 business came over to Lenovo in October of 2014. I was part of that transition. Prior to that, Lenovo acquired the notebook and PC business from IBM back in 2005. ThinkPad, as you may know, is part of Lenovo. It hit 25 years in existence this year and it's still a premier product. Lenovo has done an incredible job, No. 1 in the PC industry.
It's actually one of the best success stories in the PC industry. Everybody worried about that transition. Everybody wondered what was going to happen. But the ThinkPad brand continued and the products just kept getting better.
I was a young engineer working for the executive in Japan that was in charge of [the ThinkPad line.] It was a very interesting time. TFTs came into existence and the technologies really changed people's lives. Back in 2014, Lenovo made two acquisitions— one of Motorola, the Motorola phone, handset, mobile phones, and so on. They also bought the data center group from IBM. We really have three core businesses: the PC and clients, which is our core base business, we have mobile phones, and we have the data center.
Within the data center, we have storage, we have networking, we have server, but we also have solutions. We have hyper-conversion, a lot of software defined. We believe we are open, and we're really working to disrupt the status quo. We have nothing to lose. We're coming in looking at really software defined and how we move things forward, leveraging the capability that we have.
This show is critical as part of our solutions. Back in 2010, we worked closely with SAP to develop HANNA. We had a small development team that still sits inside of Waldorf, Germany. They worked on it, obviously SAP's code, but we were a partner since then. The first [HANNA] appliance delivered in 2011. We ship well over 7,500 appliances now.
Let's explain a little bit about what HANNA is and how it works. What made it different?
HANNA is an in-memory database. You basically bring all your data in-memory and do analytics on it in-memory, as opposed to going off the disk. That's the simple view of it.
This show, as you look around, is very professional. I go to a lot of the shows and a lot of events. I manage HPC [High Performance Computing], who's a very different clientele, but this one is actually business getting done. You see the way the tables are set up, you go out on the floor, and people are actually talking business, talking business outcomes, talking about "how can we move and transform enterprise into the digital world?" That's what's exciting about this show. It's actually around real business in all the outbreak meetings as well as all of the sessions. It's all around business.
SAP is building software solutions. Lenovo is providing a lot of the backend hardware to run and power all those solutions.
Infrastructure, yes. In our relationship with SAP, we look at it as a partner and a customer in a strategic alliance. As a customer, SAP is a big customer of ours. They run their HANNA enterprise clouds and their I.T. clouds on our platforms. They say that publicly, they run it. In addition, we run SAP. By the way, they also use our ThinkPads and everything else. We run SAP. We've got one of the largest instances in China, in Asia-Pacific, all of Asia-Pacific.
Last year at Sapphire, we won an award for customer innovation. Bringing multiple companies together created a lot of IP challenges from different disparate systems. One of them was in our cost forecasting and our vendor management. Working with SAP, we invented some new technology. Lenovo was not on HANNA at the time, but we moved over to HANNA on our platforms, development forecasting. Where is that tornado going to land? What's the threat and the need for emergency response vehicles in this area versus in other areas so that you can optimize it?
When you look at the span of what HPC is doing, safety, a good example. We call it a racing arm of Lenovo, where we can advanced technologies. You look at Formula One. ABS brakes came from Formula One. The little push button start came from racing. Airbags came from racing. All that stuff was modeled and simulated using high-performance computer systems.
It has been democratized. It's not just big companies.
Not just big companies, because the cost to compute is down. You'll see more and more cloud-based. You'll see more and more of the Googles and the Amazons, and the IBMs providing compute clouds for HPC. Just go run a batch, and go run your workload.
Since I have never bought any compute and I don't manufacture my own, could you talk a little bit about the differences between purely cloud-based solutions and then having your own system in-house?
In HPC, it's like a tool. Financial service is a good example, too. They run their grid, their credit risk analysis stuff. They want to keep that themselves. It's like the laboratory. It's like the old calculator. "I own it, it's mine." That's a big challenge.
They're afraid anybody else is going to get access to it.
Typically they like to run what's called bare metal. They don't want any software in the middle that's going to slow things down and virtualize it. They want to run right to the metal. More and more, virtualization - the shared usage model to run temporary jobs and the cost of it to go do simulations and modeling--is democratized for people to use it much more broadly. You'll start to see it in research and the dev op area, looking at how you can use the technology and the algorithms to go do things.
AI is a key example. As GPUs get faster and faster and you have the learning needed from a lot of data and building learning models, more and more of this will be cloud-based.
One of the things I've heard here at the conference is that SAP's building machine-learning and AI into their entire line, often in very small ways. It's not like "this is our AI product and it's going to solve all these problems." It's "this is our HR management product, and this specific task can be made better with machine learning, so we're going to deploy it there."
Leonardo, SAP Leonardo, is their platform. If you think about the different areas that SAP software is with Success Factors, with Ariba, with Hybris, and you look at that entire landscape. The more you can automate, the more you can speed up results and benefit yourself. The more you automate, the more you can learn, and the more you can predict. Predictive analytics is a full-circle element going forward. SAP is started in analytics. This is in their wheel-house, to go operate this way, and automate the best they can. Having those services available to be able to take advantage of those insights and then acting on those, but more importantly learning as a result, [SAP] will build very, very strong successful models.
You can feel that desire amongst the people attending the show. They sense that there's great opportunity there and they want to figure out how to harness it. One of the things we've talked about a lot on the show is that the downside of AI and of automation is that there could be job loss. There are a lot of processes that used to be done by human beings that are now going to be done much faster and much better by algorithms. Is that a problem, or is it not?
I think it's evolution. That could be said about any technology through history. I come from Buffalo, a steel mill city. When I was growing up, it was all steel mills. Lower cost manufacturing moved it all out. Thank goodness. Bad at the time, but it's reinvented. There's a lot of cities along the Rust Belt - Cleveland and other cities as well that did the same thing. You basically reinvent and become stronger and better and better. When all the outsourcing work occurred in the finance industry and even tech writing and everything, that all kind of went-
No, not tech writing! Please, not tech writing! Although it is happening...
It is happening. It kind of moved, but people reinvented themselves. Automation of those big batch areas are occurring as we speak. More importantly, it is going to help companies lower their costs but improve their results going forward. I think it's just evolution. People will continue to move.
The nature of employment is definitely shifting, but we also have to keep our eyes on the fact that overall quality of life is going up. Our healthcare outcomes are better, our businesses are more efficient. We're living in a better world day by day, in part because of these technologies.
Absolutely. I think this has changed everyone's lives, this is kind of an obvious thing. The smart phone is sometimes a pain in the butt, because you're always looking at it and you're kind of addicted to it. But some of the new applications with SAP saved about $50 million in actual transition costs that we would have had to pay to the mother companies.
More importantly, we are taking cost calculations that were in the 11-hour timeframe down to six minutes. We're taking forecasting - we ship over 100 million things a year, so we're a big company from a transactional base - from days to hours. Very, very successful. [Lenovo] won the innovation award last year at this summit, and it's how we use and partner with SAP.
You hear that story a lot, where you go out looking for a client or a customer, you wind up using them and incorporating them into your own company. Then you're a customer of theirs at the same time they're a customer of yours. It creates an interesting synergy, where you wind up understanding their platform almost as well as you know your own.
That's where it comes in. As you look at the future, we still have that same team, grown, in Waldorf - as well as back in Raleigh, North Carolina - that are working on the next generation with SAP. We've pioneered Evora with them. We're out there now selling those products. As we look to the future, we're in there on the next generation intel platform that you heard Diane talk about earlier. They're using our platforms to develop them. We're also in there with the cross-point technology. We think that will be a game-changer. It won't be available at the initial launch, but it will be available later. Again, our platforms are in there working with SAP on the development side on these new innovations with them.
Let's talk a little bit about that. When Lenovo took over the PC business, I think everybody intuitively understood, "Lenovo is going to have advantages to building PCs that IBM didn't have." Are there similar advantages in the data center that Lenovo has, that IBM just didn't have?
IBM is a great company. It goes end-to-end, very vertically integrated. Look at Lenovo as a broad company, a wide company. One of the things that we have at Lenovo is obviously scale. Shipping 100 million things a year is logistically important. Our vertical manufacturing capability, bar none, is better than any of our competitors in the industry. We're also very open. We have the permission to not be biased of any specific technology in middleware, I'll call it, operating systems. As we go to customers, it's very much like high-performance computing. When we were at IBM we could talk about Power, X86, whatever. When we went to customers, we talked about their problem. We could determine what schedule would work best, what technology would work best.
Very similar here at Lenovo. We can go into a customer and say, "Look, this environment is best suited for Azure or for VMware." We can understand and we work very closely with our strategic partners, which are VMware, Microsoft, SAP, SUSE, Red Hat, in a very non-threatening way to them. Most importantly to our customers, we're not leaning one way or another with a bias of "you need to use our storage because of X, Y, Z, or our open stack because we're on this version of Linux because we're part owners of this company or the other."
You mentioned high-performance computing a couple of times. Really it seems like some people maybe want that broken down into what exactly that means. All computing is high performance at this point, right?
All things start with high-performance computing.
People are sometimes afraid of HPC. When I was running HPC, we were really focused on the democratization of high-performance computing. If you look at it historically, you had to be a PhD. You had to be doing research in science. It was meant for the national labs. It was very expensive. I remember installing a 10 teraflop system in 2011. You'll see double, triple that coming out now.
As the cost of computing comes down dramatically, customers have a job to get done. They have insatiable demand for computing. Just getting more and more computing. If you look at it now, it's become very democratized. It's very low cost. The artificial intelligence elements of it is an adjacent space to HPC, doing a lot of analytical work. It came from batch jobs. You'd submit something, and somebody would do an analysis, and it's all parallel processes to just go attack that problem. Now it's all real-time. The faster you can get it done, the more you can get the results. It's really around time-to results and then acting on those results.
Can you give me some examples of how that process has changed? What can we can do today we weren't able to do three years ago?
In the science world, it's pretty straightforward. There's project over in Europe around the human brain project. With Lawrence Livermore, we have done things with the human heart, to model it. Drug interaction. Drug design. Cancer research. The good thing about HPC - and that's why it's so warming to Lenovo and what we do, because we are the fastest growing HPC player on the market right now. We have 99 of the top 500 [Supercomputers] list in the last go-around and some of the largest installs. We're installing some of the largest systems right now over in Barcelona supercomputer called Mare Nostrum. When you look at this, it really applies itself to good things. It applies itself to cancer research. People call this [points to phone] an appendage to me, but it does improve your quality of life. You can do things that you used to have to go sit in front of a computer to do. You can do it more casually. It changes it.
What was your first computer?
It wasn't an Apple. It was a PS1, I think. Actually I think it was probably an XT, the second generation. I was on the development team for the third generation PC.
You went straight into development?
Yeah, I went straight into development.
You had the PC, but it was also you were building the platform. You were a builder as much as a consumer.
I was a builder of it and understood it from command line prompt, DOS.
Now we've got portable computers in our pockets that are constantly connected.
And then some, connected. I think the connectivity of computing is the core, the fabric, the network. Basically that is the computer. That's what evolved and basically created the internet. Having it connect to those. Our first PCs were never connected. They were just standalone client devices. Great productivity, but they weren't connected.
Kids today have never not been connected. I'm right on the borderline. I remember when I used to have to dial up to get the internet. There was a handshake process, and I knew what it sounded like.
Yeah. I knew through the sounds whether or not it was going to go through or not because it didn't always go through.
That still happens with fax machines.
You can tell when it's falling off and you're like, "I'm not going to get the connection." Kids have no appreciation of that whatsoever. Either they have wifi signal or they don't.
I think 5G is going to change the world again. You look at 5G and you look at the capabilities of 5G, and that whole transformation is going to occur in the telephone and what's going to happen in service provider data center. It's going to be another large enablement for new platforms.
What's Lenovo's role in the deployment of 5G?
Very good. We're part of the Open Compute project, Open Compute platform, but we're also very focused and we have the workbench with an open stack distribution that we're working on with Red Hat as a partner. It's a workbench for infrastructure so that partners and clients can come and test VNFs on it. They can test their overall management. We work very closely with the telco industry in doing this. Think open stack, think about a trusted, hardened infrastructure that can support it, and then us partnering - not competing - with partners that create their own VNFs and system integration capabilities.
How confident are you that the 5G in the United States, here in Orlando, is going to be the same 5G in Barcelona, the same 5G in Shanghai?
I think they work pretty hard on it, but anything goes. Everyone likes to differentiate. The suppliers like to differentiate. I know this phone works everywhere, but some places it draws a lot more power because of the network that it's on. I think getting it to a level of standardization will be key.
Obviously WannaCry just hit the web last week. We're still feeling the fallout. The first infection managed to be somewhat contained, but I think we all know that something similar is going to come again. To me it just points out the vulnerability we have in our overall. We talked about the upside of the networked world. This is the downside of it. From a data center perspective, what do businesses need to do? What do consumers need to do?
Definitely there are going to be bad actors out there, continually doing what they're doing. They're going to exploit vulnerabilities and maybe in extreme circumstances create vulnerabilities. It's going to happen. Layers of security, obviously, are going to help in that area. Isolation is going to help in that area. I think you're seeing a lot of enterprise customers now, the number one concern is security - especially with the cloud, public cloud. For our platforms, we went through an awful lot as we moved into Lenovo. We know where it comes from, how it compounds. We build it in one physical location, which three people have access to. We do secure boot all the way up the stack. We use TPMs.
A lot of problem happens when you don't have the right patches. Some people wait for maintenance windows. What we can do is figure out how to best enable with our partners patch upgrades, without taking down systems, without having to require maintenance windows. You want to get this stuff done real-time, even though it may have been critical to do. A critical, somebody may have decided "let's do it during a maintenance window."
Yeah, and you can't necessarily wait.
You can't necessarily wait. It's a risk, Dan.
I want to get some of the questions I ask all my guests. First of all, what are you reading right now? For fun or for education.
You know, I signed up for Audible. It was a free month or something, and I still get billed every month. I can't figure out, it comes through Amazon now. I think I have like five books downloaded and five more ready. At some point you got to cut it out. Actually what I'm reading right now is a hard copy book. It's called Eat to Live.
The hard cover books are coming back. Digital sales are actually coming down, and physical book sales are tipping up. Not by a lot, but a little bit. People still like their books.
Yeah, the physical is good to have. I do have a Kindle, and I do have, obviously, this, but the physical books are better. Especially at the beach.
A lot safer that way. When we're looking into the technological future and everything that's going on, are there any technology trends that concern you and keep you up at night?
I'm excited about technology. I'm an optimist when it comes to technology. I don't come in saying that's not going to work. I'm like, "How do you figure out how to make it work?" I'm an engineer. How do you take that to market? What are you going to do? I met with a few companies here today thinking, "How do you go do that? It looks good." That's part of the challenge, because everything looks good. I'm not seeing anything that's threatening or a challenge. I think it's all pretty exciting. I think the biggest challenge is people wanting to stick to the old way and not disrupt the status quo. Clearly there's a natural evolution to when products will mature and become successful. You can be too far in front of the curve, and then if you don't have the persistence to stay with it, you kind of fall back. But I think I'm pretty optimistic.
Is there anything in particular that you think you should be really optimistic about, that's really going to transform humankind?
The cloud is a computer. That is the infrastructure, but that's kind of cliché because everybody knows that. I think individually, personally, binge-watching Netflix, Amazon Prime. I started to get into that with all my travel. The capability to stream, get access to technology or entertainment anytime, anywhere, any place on your dime, not wait a week, I think is pretty cool. Now, where will that take us comes to a level of personalization now and how we want to act and interact with technology.
We're about the same age. We came through that generation of the P2P era, where you had the internet, and then you could get pretty much all the media in the world, as long as you were willing to steal it. For a lot of that media, the only way you could get it was to steal it because the record companies wouldn't let it go, and the movie companies wouldn't make it available.
Napster was on its run. I'm not saying BitTorrent is dead now. It's still very popular; however, there are lots of legal channels for media - for music, for movies.
It's one of those things, if you can't beat them you join them. You figure out a way to monetize it. That's what the music industry and the media has done well.
It's worked out pretty well for Amazon, for Netflix, for Hulu. They're finding ways to make this work. They're using convenience and a smooth user experience to do it.
The content that's out there now is just incredible, if you look at Netflix making their own movies, their own shows. It is incredible. You have to adapt. You have to embrace it and then figure out how to go forward versus fight it. That's what they did.
Right now Lenovo's got products, notebooks and desktops, mobile phones, and the data center and the cloud. Where's this company going to be in 10 years?
Our objective is to be the most trusted partner in the data center. If you look at our heritage and our legacy in the data center and PCs, we're number one in quality. We have continuous innovation. We have 32 of the top benchmarks in our data center group. We're number 24, I think, as a company of all supply chain, not just in our domain of technology. If you look at where we're headed, it is the device to data center. We want to look at this as a seamless environment.
We also are very focused on artificial intelligence and AI. You see that our chairman pledged a $1.2 Billion dollar investment in this area. There will more and more to come on it, the ability to basically use AI and machine learning with a company like ours that touches so many customers out there, and how we can bring insights from the data center infrastructure to help customers, out to the end user's client.
It's interesting, when you talk about it you say, "We ship 100 million things." The number could go up, but also the diversity could change, too, over time.
It'll evolve. It'll evolve a great deal. We kind of look at our businesses and our PC business is kind of what we eat in the bowl today. It's our core business. You've got to protect it. Our data center, our mobile business is really the growing part of the business. Then in the future, we were making bets, we're doing innovation, we're investing in companies that we think are going to have some long-term payoff in this evolution of technology.
Excellent. If people want to find out more about the data center, find out more about where Lenovo is going, how can they get in touch with you, follow what you're doing?
I am on Twitter @BrianJConnors. That's the best place to find me. The hashtag is #LenovoDataCenters. Go to Lenovo's site, and the data center group is a clear group that's in there. A lot of consumer stuff upfront because it's a lot of what you see is our public site, but you go in and you can see everything we have in the data center. You can also follow us on Facebook and on Linkedin, and you can see everything that we've got coming out, all of our social media campaigns.
For more Fast Forward with Dan Costa, subscribe to the podcast. On iOS, download Apple's Podcasts app, search for "Fast Forward" and subscribe. On Android, download the Stitcher Radio for Podcasts app via Google Play.
This article originally appeared on PCMag.com.