Intel Corporation (INTC) Earnings Call Transcript & Summary

Okay. I think we're live and running. So first, I'd like to thank everyone for joining the Future of Technology Series. This is a brand-new event for KeyBanc Capital Markets, and today kicks off the first 5 weeks of thematic content featuring panels, fireside chats and investor meetings with over 150 public and private companies. Each week, we'll be diving into a different sector. This week, we're focusing on semiconductors and digital infrastructure with dozens of events over the next 3 days. The details and full agenda are on the website. We're lucky enough this morning to have with us Dan Rodriguez, the CVP and GM of Network Platforms Group at Intel, to host the keynote kickoff for the conference. We'll have Dan start with a high-level overview of Intel's networking strategy. Given our new virtual world, Dan spent time to prerecord this great presentation earlier this week to make sure we could get through all of the info and supporting slides. So we'll run through that first and then we'll dive into Q&A. [Operator Instructions] And I'd like to quickly read the safe harbor statement for Intel as well. Here goes the safe harbor. Today's presentations contain forward-looking statements. All statements made that are not historical facts are subject to a number of risks and uncertainties, and actual results may differ materially. Please refer to each of the participating company's most recent respective earnings releases, 10-Qs and 10-Ks for more information on the risk factors that could cause actual results to differ from those expressed or implied in the forward-looking statements in each company's presentation. So with that, Dan, welcome, and we can run into your presentation.

Thanks, Wes. It's great to be here to kick off the KeyBanc Capital Markets Future of Technology Series. I'm looking forward to sharing with you our network and edge growth strategy in this presentation and during the fireside chat afterwards. There are several key technology inflection points that are shaping the future of technology, including 5G network transformation, AI and the rise of the intelligent edge. 5G converges compute and communication technologies deliver rich new experiences and services, fundamentally changing the way we think of compute and requiring all networks to transform. The use of AI and analytics is becoming pervasive in all applications across cloud and emerging in the network and enterprises, including various on-premise customer locations. All of this, along with all the new use cases are being imagined, are fueling the emergence of the edge. New computes we've built out much closer to the user to draw powerful insights, supporting new services across a wide variety of verticals to maximize business value. And the edge needs to be built with the same benefits that are found in the cloud. Use cases need to be supported in an autonomous fashion to make quicker decisions on the fly, and there's a need for flexibility and the right economies of scale to deliver TCO and the right user experience. Intel started the network transformation journey back in the early parts of the last decade with the communication service providers and their ecosystem. Hyperscale data centers have proved out the scalability and efficiency of cloud architectures. And comm service providers are seeing the benefits in utilizing technologies that have been honed in the cloud, such as containers and virtualization, and have been leveraging Intel's expertise in these areas. We've been investing in CPUs, SOCs and virtualization technology to ensure we can run multiple comms workloads seamlessly in an optimized manner on Intel based servers. One critical moment in this journey was in 2013. In partnership with various comm service providers, we created the network functions virtualization project with ETSI, which is a European telecommunications standards body. It was a critical step in this journey. We had supported hundreds of POCs and trials. And as a result of that, nearly 20% of comm service providers were deploying some level of NFE in their network by the end of 2015 as they transformed their infrastructure into an intelligent programmable, agile and automated network. In 2017, we took another giant step forward when we put our Data Plane Processing Libraries called DPDK into the Linux community. This built a vibrant community around DPDK and is now the de facto standard for packet processing for general purpose compute. This enabled the masses to embrace and scale network transformation by moving from running workloads on dedicated hardware appliances to moving those workloads on the general purpose servers. Since then, we've seen a lot of commercial success with major comm service providers around the world, such as AT&T, Verizon, SKT, China Mobile and Telefonica, just to name a few. Furthermore, we've also been closely collaborating with new players in this space. As an example in 2019, we saw the first 100% fully virtualized cloud-native network being trialed by Rocketon, and now their LTE network has been rolled out commercially. Our network transformation strategy is being adopted, and it has enabled our networking business to grow from less than 1 billion in 2014 to over 5 billion in 2019. With 5G just around the corner and all the sorts of new use cases that are being imagined, we see the emergence of the edge. And Intel is one of the few companies in the world that has the breadth of capabilities and portfolio to enable convergence of workloads, which are absolutely critical for edge use cases. And Intel is investing in solutions ranging from IoT, network and AI to meet the same flexibility and scale commonly found in the world data centers to the edge. As we look ahead into the second half of 2020 and 2021, we will be delivering various hardware and software products in our portfolio that will be highly optimized for edge computing. A big part of our focus at the edge is also developer tools. They're critical to make our portfolio easier to consume. Additionally, they can also help democratize the edge by attracting and enabling a broad base of developers and ecosystem partners across the cloud, enterprise and communication markets. They also support numerous verticals such as industrial, retail, visual cloud and media as well as financial services and health care. This presents a large opportunity for Intel, and I will focus the rest of this talk on the edge. So why should we care about the edge? According to analyst firms, in 5 years, humans and machines will create 10x more data than we did this year. And 75% of that data will be created outside of the data center, in the physical world of factories, hospitals, retail stores, cities and many forms of video. More than half of this data we process is stored and analyzed closer to the user to deliver the right latency, bandwidth, reliability and security and privacy for a wide variety of use cases. Today, services that reside within the customer premise or go through some type of cloud. In the future, the compute on the customer premise will become much more intelligent. Furthermore, the network edge will start to deliver new compute services to enterprises and vertical markets in addition to network services. For example, most industrial factories require a high degree of data processing on site. Fast and accurate defect detection, near real-time image processing and the need for advanced security in their devices as well as industrial PCs. In order to truly deliver smart factories in the fourth industrial revolution, we need to bring digital and physical technologies together including AI, IoT, robotics, network connectivity and cloud computing. Retail is another excellent example where we see the rapid pace of technology innovations happening at the edge. Data flowing from retail stores carries important contextual information such as inventory, traffic flow, purchase frequency and real-time customer behavior insights. This data combined with sensors, computer vision and always-on connectivity are delivering real-time insights and empower retailers to deliver more immersive and engaging shopping experience. All the new use cases will push and pull the platforms and network in all sorts of directions. Some of the use cases will require both high bandwidth and low latency, and some will require higher levels of privacy and have a need to store the data locally. And many of the use cases will require 24/7 connectivity with no loss of service. There are a lot of interesting use cases for the edge. At the same time, it is complex since it's vertical and application-specific. There are so many factors that need to be managed and multiple players that need to bring it together. The various vertical markets and enterprises are focused on delivering seamless and amazing experiences for their customers. To achieve this, we as an industry need to deliver the underlying tools and ecosystem to build out the edge. Everyone has a role in [ amount ] of diversity. It's truly a vast ecosystem. Equipment providers, software providers and system integrators are driving vertical-specific innovations that are targeting multiple industries. Some players have deep expertise in these industries or different verticals and can help to pull components and ingredients together to over vertical-specific solutions very quickly. Cloud service providers are extending their cloud presence to the edge with their unique initiatives and creating a smooth cloud-to-edge developer experience. Communication service providers are actively investing to ensure their network can meet low latency and high bandwidth services. Additionally, they're also investing in building out their edge to support compute services in addition to just supporting network services from these locations. The edge is truly a massive opportunity, and it's going to take the entire ecosystem and value chain to deliver the promise of edge computing. Deploying service at the edge is complex, and it needs to comprehend several factors. For example, there are multiple edge locations across various types of customer premises and the network edge. Many of them have unique deployment environments from factory floors to cell towers to network and regional data centers to outdoor ruggedized conditions. Hence, platform solutions for the edge need to not only leverage and reuse proven cloud technologies for agility, but also require edge-native optimizations that are unique to edge use cases and deployment environments. An example of the edge-native optimization could require the network to be time-sensitive to meet the low latency requirements for industrial use cases. Additionally, multiple clusters of edge nodes need to be seamlessly managed and orchestrated in an autonomous fashion with real-time decision-making and an operational model to reduce TCO. Edge use cases are highly diverse across numerous verticals: industrial, retail, health care and others. In order to deliver optimal business value to the end customers, the solutions need to not only comprehend key drivers such as latency, bandwidth, performance, but also take into account security, privacy, data [synergy] as well as TCO. Edge solutions need to be also highly consumable with the best user experience for the end customer. Additionally, in order to enable new forms of service revenue for our customers, we also need to enable cloud and edge developers with the best experience, so the applications and service can be created and delivered at the edge. Finally, the edge applications require uncompromised performance. Different use cases have unique and demanding KPIs from security to quality of service, to latency and to bandwidth. Proper load balancing and automation as well as performance tunings should also be taken into consideration. Intel has been collaborating with the industry and ecosystem while continuing to invest in silicon and software to address these challenges. We believe this will require a convergence of key technology at the edge as this calls for a combination of compute, analytics, communications and more. Intel's diverse portfolio allows our customers to create solutions that fit the requirements and constraints of each location in a very flexible and scalable manner. From a silicon perspective, we've been investing in CPUs and acceleration technologies to enable our customers to scale their solutions across a wide range of locations and applications to support the right power, price and performance. Intel's Xeon scalable processors offer strong performance and flexibility. And they are ideally suited to enable scalable and agile 5G-ready network to deliver services from the cloud to the network core to the edge. Intel Xeon D family offers greater integration with integrated Ethernet, cryptography acceleration and compression in an SoC form factor that delivers strong performance density or performance per watt for space- and power-constrained environments such as some essential offices or on unique customer premises. Intel's core processors deliver the power and responsiveness to support many types of IoT devices including kiosks, industrial PCs and more. And our Atom SOCs are suitable for price-sensitive and power-constrained environments at the customer premise and network edge locations. Additionally, our platform-level technologies like Ethernet, FPGA, memory and edge-ready accelerators, such as Movidius, are here to help our customers support their solutions. Intel's software strategy is really to create an easy button for developers. We're offering a wide variety of open source development software to enable our customers to get to market quickly in addition to enabling them to get the most performance out of our silicon and scale across our wide and vast silicon ingredients. And at the same time, we aim to support industry standards, promote a wide and robust ecosystem and enable multi-cloud environments. We believe the critical workloads at the edge are network, media and AI, and we invest in software solutions targeting these workloads with specific recipes and development kits for these vertical markets. Aside from DPDK for networking or packet processing, we're also investing in several developer tools. OpenVINO is designed for inferencing at the edge and enabling customers to use AI in various computer vision applications. We have a large set of developers and is still growing, and customers are already deploying it in various edge use cases. From the networking side, we have OpenNESS. It abstracts out the complexity of the network for developers and exposes standard APIs for the developer community to enable all sorts of applications. We are starting to see a lot of customer traction on this developer tool. We also have Open Visual Cloud, which is an open source project that offers a set of predefined reference pipelines for various target visual cloud use cases that can be deployed in multiple locations. These reference pipelines are based on optimized open source ingredients across several core building blocks, which are used to deliver visual cloud services. Using these developer tools along with vertical-specific software, Intel is enabling the creation of recipes that are highly optimized for vertical market use cases at the edge. At the same time, we know that our customers want to tap into the broadest ecosystem for choice in their edge deployments. We are collaborating with the ecosystem to optimize scale and deploy the solutions at the different edge locations. We are also enabling them to optimize and differentiate their solutions with an open ecosystem and industry standards. To help our customers jump-start their edge development, we are partnering with solution providers and ISVs around the globe to focus on vertical-specific applications on top of Intel's platform. The phrase "It takes a Village" crystallizes at this point, and we need to do the work together with our ecosystem, our partners to convert the potential of the edge into reality. Many industries are driving innovations at the edge. As you can see, the real-world example shown on this slide range from health care to automotive to wireless networking to media delivery and the cloud gaming. I'll go through a couple of these use cases now. Audi uses Intel-based edge analytics and machine learning to automate, enhance critical quality control processes in its factories to increase the efficiency and reduce costs. For example, at one of Audi's sites, they assemble 1,000 vehicles each day with 5,000 welds in each vehicle, leading to 5 million welds that need to be inspected each day. The predictive quality control solution improves the quality control processes for the welds on these vehicles and eliminates the need to perform manual inspections, which was costly, labor-intensive and left too many unanswered questions about the quality of all cars so they couldn't inspect each one. Another example is CenturyLink. They saw the increasing demand for high-performance applications with computing capabilities at the edge where latency, bandwidth and geography were all critical considerations for the close interactions between workloads and data. They're running virtual network functions, in other words, network services and standard enterprise applications on Intel's Xeon Scalable processors from their network edge. They are also running emerging applications for retail, adaptive manufacturing as well as AI-powered video analytics. Let me dive deeper into one of the customer examples to show how Intel silicon technology and these developer tools are being used to accelerate time to market for our customers as well as to address their pain points. China Unicom is investing in a multi-cloud environment for the edge. They partnered with Intel and integrated not only our silicon, but also OpenNESS into their mobile edge computing software to see traffic to the right edge locations in order to deliver the right SLAs for unique use cases. In addition, they also used OpenVINO for use cases such as smart city, for live navigation, traffic warning, traffic scene recognition, which reduces accidents and city traffic congestion. Additionally, they've recently launched a cloud gaming application optimized using the Open Visual Cloud capabilities on a China Unicom Platform-as-a-Service offering. This multi-cloud MEC solution allows China Unicom to also run applications from cloud service providers such as Tencent as well as our own application services. This China Unicom 5G MEC solution is now in commercial deployments in 3 provinces in China with Intel as a key collaborator. It is running at the network edge locations and utilizing 5G's low latency capability to deliver all sorts of services. There are several technology inflection points that are shaping the future of technology, including 5G network transformation, AI and the rise of the intelligent edge. With all of this, we'll see the build-out of compute in many edge locations to deliver powerful insights to drive new business outcomes. And it's absolutely critical that we have a strong and robust ecosystem to pull these solutions together. And Intel is excited to have this opportunity to partner with the industry to deliver the future of the edge. Thank you very much for your time. And I look forward to talking with Wes in the fireside chat.

Okay. Well that was a great presentation. I think we're back live now. [Operator Instructions] We'll have about 15 minutes. I do know that a lot of investors are very interested in Intel's manufacturing strategy. The company did discuss the 7-nanometer delays on the call a couple of weeks ago. And given Dan's focus on networking and our broader conference focused on data center expansion, 5G and other networking trends, that's what we'll dig into today. For the first question, maybe we can touch upon it sort of in a roundabout way. And Dan, I'm just wondering how is your business impacted by the recently announced 7-nanometer delay? And is that leading edge manufacturing capability necessary for some of the products that you're delivering?

So first, Wes, it's really great to be here with you today and talking with you and the rest of the audience. So in terms of the 7-nanometer delay really, I'll talk about it really from a global perspective on our data platform group. So first, in the near term, there's no real impact as we continue to ramp 10-nanometer. We have a strong set of products coming out throughout 2020 and through 2021. And when you think about our goals, our overarching priority is to really deliver product leadership on a predictable annual cadence to meet our customers' needs. And process technology is one of those key innovation areas. We're also innovating on many other fronts, including software, architecture, memory, interconnect as well as security. And those are known as our 6 pillars of innovation. And if you think about the diverse set of workloads that will truly traverse our data-centric environment, it's critical that we draw upon many of these different innovations to be able to meet our customers' needs. Now having said that, we're of course continuing to invest to be an IDM. And with that, obviously process technology is critical, and we're going to continue to invest in process technology. But as we go forward and we look at 7-nanometer, that delay is not ideal. However, I am confident that we are going to make the right objective and pragmatic decisions and using some of the contingency plans as necessary, such as our disaggregated strategy.

Understood. Now let's dig into some of the networking pieces. In the presentation, you discussed the idea of network transformation. I was wondering if maybe you can just tell us a little bit more about why customers are focused on this? What's changing? And then you described a $65 billion opportunity and -- or I think $5 billion of revenues today. But if you can talk about the ramp in revenue or the opportunity as you expect it over the next few years, that would be helpful.

Yes. Absolutely. And Wes, if you don't mind, maybe I'll take a little step back in terms of what -- kind of what network transformation is, and then we can talk about the revenue just a bit. So first, in terms of network transformation, this is something -- this collective kind of journey is something we started about a decade ago. And it was really a direct partnership with our end customers, the comm service providers. And in that time, they came to us with kind of a simple but [indiscernible] a very challenging statement. Essentially, they wanted to move the world of networking to be able to utilize some of the technologies and tricks and tips that you commonly found in the data center, into the world of networking. So essentially what that meant is they wanted to use general purpose servers throughout their network and cloud technology such as virtualization, to deliver the right sort of scale and agility as well as flexibility into the world of networking. And if you think about the way networks were designed in that time period, they were designed with using fixed-function appliances or single-purpose appliances. So those appliances could do one job really, really well but they weren't flexible enough to be able to support multiple workloads. And it gives you an example of an industry, something that's totally different, and when you think about your PC. Obviously your PC runs multiple workloads: PowerPoint, Excel, browsing, et cetera. You wouldn't dream of having a single machine for each of those workloads, but essentially, that's the way the network was laid out. So with that in mind, they came to us and they essentially want us to bring data center economics into the world of networks to drive overall TDM, reduce R&D and increase their overall TCO.

Okay. That's a very helpful description. So it sounds like in terms of the product road map that you would have a wide range of products to fill some of the simple solutions, but actually maybe you sell a lot of server-like products. [indiscernible] processors, is that correct?

Absolutely. So when we think about what we're developing to be able to meet that need, essentially what we're developing is a slate of products, including our Xeon Scalable processor, which is a critical product for not only network but also edge. When you think about markets such as the wireless core, they're going to truly require a lot of service flexibility in the future as we head into 5G. So Xeon Scalable processor has been in mainstay in LTE networks and we expect that to accelerate. In addition to the Xeon Scalable processor, we have a slate of SoCs ranging from our Xeon D, which is a Xeon in a SOC form factor, to be able to work in more power-constrained environments to our Atom class SoCs for power and price to a host of different accelerators as well as platform level ingredients. And if you think about what that's offered on Intel is it enabled us to significantly grow in this market. And going back just one short year ago in 2019, our business did hit roughly $5 billion.

Okay. You also mentioned in part of the discussion around the edge and the presentation, lower total cost of ownership and I'm just wondering how Intel is providing that to customers.

Essentially, when you think about the lower cost, TCO, there's a few things that are really driving that. The first is that when you move to general purpose servers, you have the opportunity to really draw upon a robust set of ecosystem players. So essentially you can draw upon a robust set of OEMs to deliver an ODMs deliver the hardware, but also you can draw upon many different software vendors, whether that's supporting a network service or what we call a virtual network function or perhaps a new compute service. And with that separate software and separate hardware, plus some of the automation frameworks that are found in the cloud as well as IT data centers, you combine all that and you can enhance you [ volatility ] of your network, but also reduce the overall total cost of ownership.

Okay. And I want to come back to the software side in a minute. You also mentioned a ramp-up in edge solutions in second half '20, first half '21. I was just wondering if you could run through some of these. Were those examples you gave in the presentation? Or are there some additional products or opportunities you could discuss?

Absolutely. So when you think about some of the edge solutions, again, I would think about a few different things. So one is obviously we have that breadth of CPUs to split a wide variety of environmental conditions. I mentioned -- I think in the video everything from factory floors that you can also imagine a base station out in either a desert or some sort of mountain climate where you have all sorts of different environmental factors. So it's critical that our ingredients support all of that. But in addition to this, we're also investing in specific set of software. And oftentimes, what we're doing with that software is we're creating vertical-specific recipes. So it could be something for industrial, it could be retail, it could be something for a smart city. And we package all that up in a way that's easy for our customers to consume so they can get to market very, very quickly.

Okay. That makes sense. I also wanted to touch on -- there was a recent announcement with VMware. And I was wondering if you could discuss that and how it ties into some of your efforts.

Yes. When you think about the announcement with VMware, this announcement was really centered on the radio access network or the base station market as it's commonly referred. And when we think about the RAN, we think about comm service providers really driving all sorts of different requirements. We're seeing the RAN need to support multiple different vertical markets in addition to traditional kind of public networks, if you will. As a result of this, we do see a traditional RAN as well as a virtualized RAN that utilizes Intel's flex RAN solution that's really coexisting seamlessly. And from a virtual RAN standpoint, the idea of being there is to be able to drive the same sort of scale and agility that you're now starting to see in the core network, but also apply it to the RAN. And with this strategy, we're starting to see a lot of momentum behind it. We're starting to see comm service providers trial these solutions as well as some even start to deploy them. And with respect to the specific collaboration with VMware, the idea being there is to really simplify the integration of virtualized rent. Essentially, we're working together to drive integration and reduce the steps it takes to deploy [ VRAN ] solutions at scale in the marketplace.

Okay. That's helpful. Let's come back to software for a minute. I'm just wondering if you could tell us a little bit more about Intel's investment in software, how that's helping Intel and why you think it's necessary.

So when you think about our software investments, you can really think about it from a couple lenses, if you will. First, from a software perspective, we are investing to ensure our customers can get the right performance out of our wide variety of silicon platforms. Additionally, the way we do deliver our software, we try to use industry standards where possible, and we try to promote on the use and be able to draw upon, again, an easily -- again, the many silicon ingredients we have in our platform and be able to seamlessly scale across those. And that's possible because we do have architectural consistency across our CPU families. And then the second reason we invest significantly in software is that we fully believe that to be able to drive the innovation cycle up in the world of networking, we need to create a robust and really open ecosystem. So because of that, we're not only partnering directly with ISVs, but we're also partnering directly with many different standards organizations or consortiums such as the Linux Foundation, TIP or ONF to be able to drive this wide and robust ecosystem.

Okay. That's helpful. So it sounds like the software is a key sort of partner, if you will, to the hardware to create those overall [ visibility and ] cost of ownership of product. Let's spend a minute on 5G. It's a big -- the ramp is happening now. It's a big topic for Intel. Where is Intel winning in 5G?

Yes. So when you think about 5G, 5G is really obviously about delivering all sorts of new bandwidth as well as capacity on the network. And people immediately think about consumers, but with 5G, it's also about delivering all sorts of new business outcomes or business services, essentially creating new services to drive new revenue streams or just enhance customers' experiences. So with that tailwind, it's really providing a ton of momentum for our overall network transformation strategy. And if you think about the core network for a second there, essentially in 2019, you saw roughly 40% of all deployments in the core network being virtualized network server. Furthermore, in 2020, that number is expected to be 50%. And by 2024, that number is expected to be over 80%, again with 5G kind of -- 5G as a tailwind. And then if you think about that and then you think about the overall market just for virtualized network servers, that really cuts across core network, enterprise networks and many other types of networks. We're seeing that part of the network grow at roughly a 20% clip, whereas fixed-function appliances, which I explained earlier, are only growing at roughly 1%. Now in addition to this, when we think about one other key aspect of 5G, it's obviously the radio access network. And in the radio access network, you go back to just a few short years ago in 2014, we had no market segment share within that segment. Since that time, we've invested and we acquired. We made 2 acquisitions. We've done organic investments. And as a result of this and the strong product lineup that we have, we're anticipating and projecting a 40% share in 2021.

All right. That's helpful. I know we don't have a lot of time left. I did want to hit a couple more topics real quickly. I asked a little bit earlier about the Xeon opportunity. I'm just wondering if -- as we think about the edge and networking, can you give us an idea of the mix between the various products, SoCs, ASIC, FPGA, Xeon in your group?

Yes. Absolutely. And we -- obviously this breadth of products really does give us a really strong competitive differentiation in the marketplace. And we think about Xeon as compared to the other products, the way I see the business now and going forward is that we do expect a very healthy mix of Xeons within the mix of our overall volume and revenue going forward. And then when you think about the SoCs the FPGAs and ASICS, some of the newer product lines, I do expect them to really continue to gain share for us. But even with that gaining share, I do expect Xeon to again be a healthy mix of the business. And then if you -- that's really on the network side. When you think about the edge, we're going to still see those same kind of mix of processors in there. But additionally, we also sell the Core processor in there that's commonly found in laptops as well as PCs. But we're working with that team to ensure that the Core product line can meet all sorts of use cases really in the world of IoT such as retail and industrial.

So would it be fair to characterize the Xeon as the main sort of profit driver of this growth opportunity?

Absolutely.

Okay. That's helpful. I'm also wondering about competition. Who should we think about as your competition in the networking space? And how does that change maybe as we move to the edge?

When you think about competition, I think you could probably guess who some of the competitors are. But it's kind of ARM-based competition and another x86 SOC competitor. In addition, as you go out into the edge, you're starting to see some AI accelerators. So when you think about what Intel does at the network and the edge, our position is quite unique. If you think about the robust product portfolio that we have, again starting with Xeon, which provides lots of performance and flexibility for core network, enterprise network and other networks to start it off, but then we also have our Xeon D product line. That's great for performance, performance per watt type solutions or power-constrained environments, our Atom class processors for power and price-constrained segments and then a host of platform-level ingredients. When you think about one of those ingredients just as an example, our Ethernet controllers, really works seamlessly with our Xeon processing for networking, essentially efficiently passing packets up to the processor and then back out through the network. So you can see that as we optimize our platforms, we consider our Ethernet controller as well as our Xeon. So that gives us a pretty good base to compete on in the marketplace. And then if you couple all that with our software strategy where we're truly investing to really enable our customers to tap the most performance out of our processors and platform-level ingredients as well as the ability to scale across the wide variety of silicon platforms we offer to be able to address multiple different workloads and locations, it gives us a lot of strength and a lot of confidence as we head into both network and edge.

That is very helpful. Dan, I wish we could talk about this much longer. This is very interesting to me, but we're running up against the end of our time. So I would love to thank you for providing the keynote and the presentation and going through the Q&A this morning. It was really informative. And we really appreciate your participation.

Yes. Thank you so much for the time, Wes. It was a great chatting with you. Thank you.

Thanks.