Intel’s Lisa Davis On How Xeon 'Cascade Lake' Will Drive A New Wave Of Digital Transformation
As a Three-Time CIO, She Has The Inside View
Lisa Davis knows a thing or two about digital transformation.
As a three-time CIO, Davis has had an inside view of the challenges that organizations face as they go through technology modernization projects. Now as general manager of Intel's digital transformation business within the Data Center Group, she's using her 30-year career in tech to help enterprises and governments understand how the chipmaker can guide them through their digital transformations.
"Digital transformation is about leveraging technology to build digital services and technologies for customers to improve their customer experience, right?" Davis said in a recent interview with CRN at the Intel Partner Connect conference in Denver. "So it's never about technology for technology's sake."
[Related: New AI-Friendly Intel Xeon Scalable Processor Packs 30 Percent Price-Performance Punch]
With Intel launching its largest portfolio yet of data center products this week, Davis' role is critical: She must communicate to organizations the business value of Intel's new products, including the second-generation Xeon Scalable processors, and how they fit within their digital transformation strategies.
For the new Xeon Scalable (previously known as Cascade Lake) processors, she makes it sound like a piece of cake.
"Cascade Lake is all about our new memory technology always like to simplify it for our customers: the bigger the data set I can put next to the CPU, the bigger the problems I can solve," she said. "And to me, that, from an analytics and AI perspective, of driving those workloads and driving insights from those data sets for our customers, is a huge value point to me."
What follows is an interview with Davis, in which the Intel executive discusses the chipmaker's approach to digital transformation, how that plays out in the real world with customers and how the company incorporates new products like Optane DC persistent memory into IT modernization projects.
For you, what does digital transformation mean?
Digital transformation is about leveraging technology to build digital services and technologies for customers to improve their customer experience. So it's never about technology for technology's sake. Digital transformation is about enabling that business to differentiate themselves to, frankly, survive and compete in the market today.
So it's all about using incredible technologies that we have at Intel. And many times, when I sit down with a customer: first, understand the business problem that they're trying to solve, which is the first thing that's most important. We like to call it Intel customer obsession: listen more, understand what the customer is trying to solve, and then look at our broad portfolio of technologies, whether it's CPU, whether it's memory, whether it's accelerators, whether it's network, and figure out how do we leverage those technologies to solve their business problem or, frankly, differentiate that business or mission set when it comes to government.
How would you describe Intel's approach to digital transformation projects?
Most of the time digital transformation — it's not a one-stop shop. It's a journey. And many times, when I talk to customers about digital transformation here, it is a strategy that frankly should be connected to the business strategy, which is ultimately connected to a data strategy. Why? Because at the end of the day, with this explosion in data that we're all seeing in companies to harness data and drive insights out of the data, we want to help a company connect the three of them. First of all, what is the business trying to do? Second of all, how does your IT strategy align to the business? Because most of it's going to be probably about getting more agile, getting more modern, getting more responsive, knowing where the data is setting and serving up that data so that the business can leverage it and drive insights.
It's connecting those three strategies together and then it's really understanding, "well what's [the] current state, right?" And having been a three-time CIO, having done this journey many times, in many different sectors, what's the current state of my infrastructure? Where is my data today? Do I have data, for instance, sitting at the edge, which is so important now?
And what does that look like?
I think the data point is more than 50 percent of data by 2020 will be created at the edge versus the data center or the cloud, right? So what data sets do I set at the edge? What data sets do I have within my data center? Where can I burst to the cloud to drive TCO [total cost of ownership] and better agility? And how do I create this distributed ecosystem that allows me to move workloads and data to ultimately harness data and drive insights for that business?
So every enterprise, they are at different places. I would say every enterprise is dealing with legacy aging infrastructure. And one of the things I like to talk about when I sit down with our customers is: first of all, it's really hard to be responsive to the business if you got aging infrastructure. Not only are you slower, you have security concerns that can't be addressed, and it's ultimately costing you more money.
So the basics to me is: when you have enterprise software vendors, and we all have them — there's SAP, there's Oracle — when there are end-of-life software licenses, it's a great opportunity to upgrade your hardware. And why do I want to upgrade my Hardware? I get better performance. I get better TCO. I get better security capabilities in my hardware and, ultimately, it makes it easier for me to manage that hardware, right? So it is about hardware, certainly in the complexity of the data center environments that we have today. So it's depending on what that aging infrastructure looks like in terms of digital transformation. It's trying to help that customer understand where the data sets are if they were going to do an SAP Hana implementation and work with our partners in SAP to do that.
What customers has Intel has worked with for digital transformation projects?
Here's a perfect example in digital transformation. Kroger today is a grocery store, right? And Kroger today has cameras in the store, and most of those cameras were probably for physical security, right? People stealing things. But now Kroger is on their transformation journey, and they came to Intel for help, and we're working with them. And they said, "you know what, I'm going to take those cameras, and I'm going to double the amount of my cameras over the next six months, Lisa, and by the end of next year, I'm going to quadruple those cameras in the store." We're like, "what are you doing with the cameras?"
They're going to put the cameras in the aisle, because now they're going to become a media company. Not only are they going to leverage those data sets for facial recognition, body movements, how I'm reaching for products, what are my facial expressions when I'm looking at products. But they're also going to leverage those media services as advertising. So I'm walking down the aisle, and Coca-Cola could buy services from Kroger to do advertising, right?
So part of that digital transformation we're working on with Kroger, as an example, is how do we build an edge-to-data center infrastructure, so they can take that data coming in from those video feeds? How much of that data needs to be computed? How much needs to be analyzed? Where do I store that data? Which is what Intel is all about: compute, storage and network. How much of the data needs to come back and be stored in my data center? And which part of the data can go to the cloud, right? Because I want to do something else with it. That's an example of how my organization, with our partners across Intel, really engage with our customers to solve some of these problems — and the problems are only becoming more complex.
What kind of kind of partners are you typically working with?
One of the phenomenal things about Intel is our incredible partner ecosystem. So I work with all partners. It's a distributed ecosystem. So obviously when we think of a customer or an IT enterprise, all of those key enterprise application [vendors] that I mentioned: Microsoft, SAP, Oracle, SAS, our ISV partners. We work very closely with our OEM Partners: HPE, Dell, Cisco, Lenovo. We work very closely with our [systems integrator] partners. Why? Because enterprises leverage those system integrators to help them integrate and implement the technology into those environments.
And then we work also with our [cloud service providers] because now — interesting for my business, which was primarily focused on on-prem or hybrid cloud and around analytics — you now have offerings like Amazon [Web Services'] Outposts, which are targeted to on-prem workloads, right?
So we work with everyone. And I think that is one of the strengths of Intel: how we enable our partner ecosystem to be successful on the technologies that we bring to market through our training programs, etc. How do we work not only in selling to those partners, but also how do we build solutions together that can differentiate and solve problems in the market for both of our end-user customers?
How are you helping partners go to market with solutions?
One of the things we launched last year was our Intel Select Solutions. So this is a great example of many of these enterprise applications that our customers use today. So if I have to upgrade to Microsoft SQL Server 2019 or VMware's vSAN product or vSphere product or SAP's product, we have a reference architecture, proven benchmarks that we can provide to our partners, so that they can leverage our technology easier, but probably most importantly from a customer perspective. The whole point is to get easier and faster adoption, to get our enterprises up and running on the latest technology faster. So we do that for our analytics and [artificial intelligence] workloads.
We're looking at developing solutions now from the edge back to the data center and trying to figure out, what are those horizontal kind of infrastructure capabilities, whether it's compute, network, storage sitting at the edge and how would that integrate into a typical enterprise?
So we have a wide variety of solutions that we provide, because in my organization, we do a lot of "sell with" where we're sitting down together with [a company like] Dell and their ready-made solutions or a Lenovo or a Cisco or an SAP or an Oracle and say, "okay, what are your customers asking for? What are you interested in doing for your customers?" And part of my job in terms of being the general manager is to find those win-wins. Where is it a win for our partner? And where is it a win for Intel? And ultimately, where's the win for the customer?
How many Intel Select Solutions do you have now?
I think we have, and we'll have to get exact number, but I think we have 30 solutions today across a variety of workloads: infrastructure workloads as well as analytics and AI workloads. And they're ready today with many of our OEM partners and SI partners and some of our ODM partners. And then what we're looking at now is what more can we do from a solutions focus? Because many times if you look at the ones that I named — Microsoft, VMware — these are all pretty standard.
What we're looking to do and where a lot of our focus now is: what are some more innovative things that we can build that aren't standard things that an enterprise deals with that solve some of these unique, complex problems as we look at workload mobility, workload orchestration between edge to data center to cloud, right? And that's where I'm probably most excited about, that analytics and AI area and then edge to what I call core — because to me, the core is either data center or cloud. All cloud is good cloud, whether that's sitting on-prem, private cloud, public cloud. How do I manage that workload mobility and portability? Which is not easy to do today. Folks say they can do it. They really can't do it. It's very hard. How do we drive that seamless orchestration for our customers? And that's really an area where I think Intel can play a pivotal role with many of our solution partners.
When did the edge become a focus for your organization?
Well, Intel's been a leader in 5G [for] years. We have led in 5G, continue to lead in 5G, and we partner with our network group as we look at solutions from edge back to data center. So I think in my organization from a digital transformation standpoint, you really had enterprises start talking about edge compute — running workloads like video surveillance workloads and then figuring out how that workload plays into my broader data strategies — probably in the last year and a half.
Even if you look at Gartner research or Forrester research, you know, it's interesting. I had a recent meeting with Gartner, and even Gartner is still developing their perspective on how the edge will continue to develop, because it can be customized by verticals right? Whether I'm in manufacturing, whether I'm in health care, whether I'm in retail.
And what Intel is doing with its partners now is: in all of those different verticals in terms of edge, storage, network and compute, what are the horizontal opportunities in terms of infrastructure that we can find and build with our partners? And it's a work in progress today — learning from many of the use cases of like a Kroger, like an Exxon Mobil, like an American Airlines, who are solving these types of problems of how Intel with our portfolio of capabilities can solve some of these problems more from a horizontal layer that then is applicable to multiple different verticals.
I think from a comms provider and from an IoT perspective, that's been going on the last five years. But from an IT enterprise perspective and the majority [of companies] understanding, "oh, I now have a capability because of compute, network and storage of being able to do that at the edge. What workloads can I run at the edge and leverage to drive differentiation for my customers and my business?" That's rather new, I would say in the last year and a half.
To what extent are you looking at Intel's future roadmap, whether it's Xeon, FPGAs or Optane, and thinking about how that fits within digital transformation initiatives?
First of all, take our [Xeon] Skylake platform that was launched last year as an example. What was the value of Skylake in a digital transformation? First, it's really hard to be agile and responsive if I have five-, seven-, 10-year-old hardware sitting in my data center. And it's more expensive. So the beauty of our SkyLake platform and why the platforms are important is because what used to take me four VMs on my aging equipment I now can do with one. I got a 30 percent performance hit, I had security capabilities that never existed — and we know that cybersecurity is the No. 1 issue for all enterprises.
So I have a platform that is designed and tuned to run the workloads that enterprises are dealing with today. One of the things that Intel does is: we do incredible benchmarking and performance. So when we build our platforms, [we work] to understand what are those workloads that our customers are addressing today, whether [they are] mission-critical workloads, AI and analytics workloads or what I call commodity workloads, such as email, devtest, those types of things. And when we build our platforms, we take not only workload considerations, but input from our key OEM partners and CSP partners and essentially build a platform that's going to help our customers in their digital transformation journey. So our platforms are designed for the workloads of today, tuned and benchmarked to perform and drive better capability for our customers. That's the beauty of what they do.
What separates Intel from other companies moving ahead with digital transformation initiatives?
I think what differentiates Intel, and I'm going to speak from a customer [perspective] as well, is, I keep referring to it as a portfolio of technologies. What do I mean by that? Intel has been going through a transformation of being a PC-centric company to a data center company. Why? Because of the explosion of data, and it's all about storing more data, moving more data, and computing more data, which is absolutely critical for today, right? If you can't do those three things, you will likely go out of business, bottom line. Or Amazon will do the business, right? I mean, that's the market. It's a dog-eat-dog world out there today. So it's about storing, moving, computing data across our portfolio of capabilities: network, memory, accelerators, compute.
So if I have a particular workload or problem set, that [involves] sitting down with the customer, understanding what they're trying to solve and say, "you know what, this is a great opportunity to put an FPGA in there." Right? "And this is what that capability would drive. Oh, you're trying to solve a data issue, and you have X terabytes of data. You're running SAP Hana. Do you understand the value of what our Optane DC persistent memory will do for you?" When I used to run and deploy SAP Hana on the Intel IT side, taking down an SAP Hana implementation, the business guys and gals were like, "you can't take down the supply chain because you [will] disrupt my operations, day-to-day operations." So taking down SAP Hana was a weekend ordeal. It was a nightmare, right? Now with Optane DC persistent memory, minutes! I mean, it's a game-changer.
This is the innovation that I think Intel brings to bear and the technologies that we have. Once we understand what that customer is [trying to solve], understanding the breadth of capabilities and technologies we have, we can then bring them to the customer, so that they can compete and differentiate themselves in the market. What is that experience they're going to bring to their customers that no one else can.
What's the digital transformation pitch for Intel's second-generation Xeon Scalable processors?
[Second-generation Xeon Scalable] is all about our new memory technology [Optane DC persistent memory], right? I always like to simplify it for our customers: the bigger the data set I can put next to the CPU, the bigger the problems I can solve. And to me, that, from an analytics and AI perspective, of driving those workloads and driving insights from those data sets for our customers, is a huge value point to me. That is really what [second-generation Xeon Scalable] is awesome about.
That is probably the most succinct explanation I've heard for Cascade Lake and Optane.
Well, having been the customer, at Intel, we can get very geeky in terms of how we explain things, so I always try to explain things so the customer can understand, "why do I need this? What is the value of the technology?" And frankly, there's been so many advances in memory alone, customers have a hard time understanding, "well, what do I use what, when?" And that's part of our job and our partners' job: is to be able to explain what all of these new technologies [are], and that's why I said to you, it's complex. With all that these new technologies that we have, what is the applicability to solving the business problems as we engage with our partners to help you?