Statements in this presentation that refer to future plans or expectations are forward-looking statements. These statements are based on current expectations and involve many risks and uncertainties that could cause actual results to differ materially from those expressed or implied in such statements.
Welcome to the Big Apple, the city that never sleeps, a cultural melting pot, and an epicenter of possibilities. It's the perfect location to celebrate the dawn of a new era of technology, one that will bring AI everywhere.
AI is everywhere, and it's changing our world for the better. Let me show you. AI is transforming medicine, analyzing millions of symptoms and procedures, helping doctors see more, so they can make the right call for every patient. AI gives us faster insights into our world, unlocking new horizons, helping us be good stewards of our planet.
Increible.
AI is expanding opportunities for everyone, helping us to learn, develop, and create like never before. The more we grow, the better we can manage our resources. AI gives us access to an unbelievable amount of knowledge, helping us plan, design, and build smarter. With no limit on where we can go. It starts here, New York City, known for its innovation, creativity, and making history. So are we. Like this great city, AI means opportunity just waiting for us. AI is everywhere, and the future starts right now.
Hello, and welcome. Today, our biggest launch of the year, and here in the Big Apple is the place to do that launch. Welcome, all of those who are here live in New York City, the chaos and joy of New York at Christmas time. We're thrilled today for Intel, for our partners, and our customers. But we wouldn't be here without the dedication and hard work of Intel employees who are joining us worldwide, celebrating our biggest launch of the year. As we think about this period of time that we're in, wow, what a year, right? All the impacts of, you know, the geopolitical environment, the business environment, and the economic cycle that we've been in, and, hey, we've had a little bit of technology as well.
And I don't know if you heard about this thing called AI, but, you know, boy, this year, this decade, and hey, team, welcome. We're ready for a launch, right? Yeah. Woo! Hey. Wow, what a year it's been. And in the world of technology, this, you know, role of computing is getting more and more present. It's imbuing autonomy and agency into everything that we do, and it's becoming part of the ambient environment in which we live. And every piece of AI built on silicon. So welcome to the siliconomy. You know, this idea that every piece of the economy of our lives is becoming more dependent on digital. Everything digital runs on semiconductors. You know, and this idea of the siliconomy, how essential it's become, God's gift of this magic little crystalline, the most populous piece of element on Earth, fueling every aspect of our environment.
You know, today, about 25% of our economy is built on silicon by the end of the decade. Extraordinary, and I think that understates the role that AI is gonna have, and maybe a full third of the entire GDP based on silicon. And, you know, here we are, and we see this driving force of silicon for a $1 trillion semiconductor market by the end of the decade. And the role of AI, just making it go faster into the future. And, you know, in a baseball analogy, we're in the early innings. Well, no, maybe we're still in warm-up. No, maybe we're still in the preseason of the impact that it's going to have. But when we think about artificial intelligence, hmm, something over there, right?
And something we may not understand or may not control, and, you know, we think it's a disservice to think about it that way. Instead, maybe augmented intelligence, how we integrate it into our human lives, into human intelligence, how we make it part of, with us in everything that we do. And we think bringing this value into the human experience is the opportunity for AI, augmented intelligence. My family, we have a, you know, disorder. We lose our hearing at an early age. My father was almost deaf when he passed away, and my AI-enhanced Starkey hearing aids are making me better. It's not something over there. It's right here, augmenting my human experience. And we think this paradigm shift, how humans and technology come together, are gonna be powerfully enabled by AI.
That hearkens us back to the Intel vision, that we're gonna work on technology that improves the lives of every human on Earth. And when we think about technology, is it good or bad? Hmm, it's mostly neutral, and it's our job to shape it as a force for good, making it everywhere, making it responsible, making it safe, secure, trusted, well-engineered, regulated appropriately. And that's what we're about as a company, enabling that experience for AI everywhere. And effectively, that's why we're here today, the AI Everywhere event, and how we're driving it into every aspect of the applications, but also every device in the data center, the cloud, the edge, the PC as well.
As we think about that, if we've seen high-performance computing, this, you know, solving these hard problems, and now AI and high-performance computing are coming together in a powerful way to enable this next generation of computing: faster memory, faster networking, larger memory capabilities, computing capabilities. When we think about this, we also have to do it in a sustainable way. Are we gonna dedicate a third, a half of all of the Earth's energy to these computing technologies? No, they must be sustainable as well. Compute efficiency, power performance, capable. With that, we're working together with research on how this machine learning and these new platforms truly solve global issues at scale. That's what we think about with the AI continuum, bringing AI into every platform and every experience. When you think about why is Intel gonna be the company to do that?
We're scale. We build world-leading technology in volume. That's what we do. But we also do it with open ecosystems, software that makes it accessible to everybody, and we're end-to-end. We're uniquely positioned to be this AI everywhere, seamlessly integrating AI into every platform that we do. We're doing that with Xeon, with Gen five today, right? We're saying, "No, you don't have to build new data centers. You don't have to stand up new networks, new management, new security. No, we're gonna infuse it into every data center that's built on Xeon," and that's what we're announcing today with Gen five. Same applications. You don't need to forklift those applications. No, we're just gonna build it in and enhance what you already do with new AI capabilities. But we're also seeing that the world is moving from high-end training, saying: How do we inference everywhere?
A few people create weather models, lots of people use them. The same is gonna happen with AI. A few people train, lots of people inference, and they're gonna do that in their data centers, in their enterprise applications, and at the client and edge. And that's what we think about with these three laws. Why is it gonna happen at the edge? You know, first is the laws of economics. My device versus a rented cloud service I pay for, economics. You know, second is physics. Round trip to the cloud, or do it right here immediately. And finally, the laws of the land, my data. You know, all the geographies that we operate in, we're gonna operate locally on models on the edge and on the device.
As this becomes increasingly important to the future, AI will dominate the edge workload, and that's what we think about with the AI PC. You know, and we've been seeing this excitement of generative AI, the star of the show for 2023, but we think 2024 marks the AI PC, and that will be the star of the show in this coming year. You know, unleashing this power for every person, every use case, every location in the future, and we're gonna take these large language models and follow those three laws and put them into every device, every PC for the future. You know, and with that, it's gonna drive a revolution in the applications. You know, hundreds of new applications, new communications, new form factors. And I liken this, and I like to say it's like a Centrino moment.
Now, for those of you who don't remember Centrino, you know, hey, you know, I helped to create Wi-Fi. And what happened when we created Wi-Fi? Pretty much nothing for three years. And then Centrino, a platform that drove in volume into the industry, and all of a sudden, every coffee shop, every hotel room, every business needed to become connectivity-, internet-enabled. New form factors started to emerge as a result. And we think of the AI PC as that kind of moment, driving this next generation of the platform experience. Andy Grove called the PC the ultimate Darwinian device. The next major evolutionary step of that device is underway today. It's called the AI PC.
As we think about this exciting year of innovation, and Intel is going through our transformation journey, and we laid out this tremendously audacious goal as we, you know, began this journey as well. Five nodes in four years. We're gonna do a decade of semiconductor work in just four years? What do you think? That's pretty crazy. On track, baby. And over here we have Intel 7, now over 100 million devices shipping, right? Intel 4, with today's Core Ultra launch in volume, millions units this year. Intel 3 going into production next year with our server launches. Intel 20A, you know, and as I said on our earnings call, the next generation transistor, power delivery, Picasso, Rembrandt, work of art. And the finish of this, 18A, bringing this all together and finishing our journey, being production ready at the end of next year.
First products going into fab at the start of the year. Now, hey, at the holidays, what do you do? You all gather around the tree, and you take a family photo. So kids, Papa's here. That's a family photo to be very proud of. You know, and as we think about this idea, you know, it wouldn't be the, you know, completion of an event like this without hearing from our friends, families, those in the industry who join us, you know, supporting this launch today, and, you know, this driving force of the coming decades. You know, and they join with us because what's Intel's purpose? AI everywhere. Technology as a force for good, improving the lives of every person on the planet. You know, and as 2023 started with cool technology, we're finishing the year with even cooler technology.
Let's hear from some of my friends and colleagues for decades right now.
Hi, Pat. It's great to be with you. The world is witnessing a transformative era, with AI and GenAI driving the latest wave of innovation. As the industry leader in server, storage, and PCs, we're in the middle of these conversations with customers who recognize the opportunity AI brings to their data.
HP and Intel have enjoyed a strategic partnership and driven critical innovations in computing for our joint customers for more than 30 years. Together, we enable the possibilities of tomorrow, and increasingly, that includes leveraging the power of AI.
The partnership between HP and Intel has led to decades of pioneering technology. I am very proud to say that HP has the widest range of client products across segments based on Intel Core Ultra. At HP, we are reimagining what a PC is and what it can do. The work Intel is doing is a game changer that will help us as we create this entirely new category of devices. Together, we will personalize the way people everywhere live, work, and play like never before. We are going to make 2024 the year of the AI PC.
Lenovo is proud to be at the vanguard of AI computing with Intel. The new Intel Core Ultra combines strong CPU performance with an AI-specific engine to handle complex AI workloads, like video editing and rendering. We are working together to deliver the world's first laptops with Intel Core Ultra, the new Lenovo IdeaPad Pro 5i and ThinkPad X1 Carbon.
With Intel's strong product innovation, combined with Supermicro's extensive building block portfolio of optimum server solutions, we can accelerate almost every data center workload.
There is no question we are in the early days of a massive platform shift that will fundamentally transform productivity of every individual, organization, and industry. At Microsoft, we are committed to helping people and organizations adapt and thrive to this new age of AI. That's one reason our long-standing partnership with Intel is so important. Over the years, we have brought Intel's innovations together with the power of our tech stack across the cloud and the edge, and together, we are building on these investments to help drive the next level of value for our shared customers. From our work on Confidential Computing, to bringing the latest Intel Xeon CPUs with Advanced Matrix Extensions to our cloud, to our collaboration on this new era of AI PC, starting with Copilot on Core Ultra, we are early shipping now.
Our future has never been brighter.
We can't wait to share more with our customers beginning early next year.
The future is now, and it's powered by AI.
I look forward to all that we will achieve together in the years ahead.
Congratulations to the Intel team for leading the industry into this new, exciting AI PC era.
Thank you to our friends and partners, and the work that we're underway with. On that note, to tell us more about what's possible, please join me in welcoming Michelle to the stage. Tell us about the AI PC.
Awesome. Thanks, Pat. Good morning, everyone. As you heard, AI is a huge inflection point, and we believe that computers infused with AI will permanently change the PC market and fundamentally reshape how we interact with our computer. Over the last five years, Intel has nurtured this intersection of AI and PCs with a steady drumbeat of new machine learning, inferencing, computer vision capabilities, all within our client portfolio. It was really to prepare for today or this moment, putting powerful AI tools into the hands of millions. The power of AI is absolutely going to grow the PC industry. We already see it driving innovation, and according to Boston Consulting Group, AI PCs are expected to comprise 80% of the PC market by 2028. All of this is driven by the seismic performance and experience improvements that we all know AI can offer.
To me, the benefit could not be more clear. Humanity is going to save tens of millions of hours. The question is, what are you gonna do with all that extra time? You might ask how. Well, email threads, creating graphics, crunching your data, completing your travel itinerary, all of these have the opportunity to make you more productive, but they have to be personalized, and they have to be useful. Over the next two years, Intel is committed to ship over 100 million client processors with dedicated AI. That's up to 5x more than all of our competitors combined. As I said, AI is a market changer, and we've been in this CPU or processor business for over 40 years, and there are some unshakable truths. CPU cores matter, graphics matter, and power efficiency matters. Nailing these fundamentals enriches the entire PC experience.
It's not just about one item. Yes, AI is absolutely a new critical dimension, but for me, it's a yes and statement. Yes, Intel can deliver a truly exceptional processor, and we can do it with a definitive AI experience, too. One processor that delivers on every front: CPU, graphics, power, battery life, and AI to meet your needs. That's our vision for the perfect processor, and boy, do we have the technology to really nail it. Today, I'm excited to announce the launch of the Intel Core Ultra. It's beautiful, isn't it? It's the fastest processor for ultra-thin notebooks, and I'm so excited because we are smashing it out of the park on conventional application performance. An example, in Adobe Premiere Pro, for example, you wanna look at our competitors. We're 40% better than our competitors with Intel Adobe Pro.
We've radically changed our energy efficiency. Compared to previous generations, it uses 40% less power in day-to-day activity. This is Intel's most sustainable CPU, and as Pat said, "That's good for the planet." On top of that, it has a world-class GPU, too. So imagine a laptop under three pounds that you can play games on, and it pulls double duty as its own powerful AI GPU processor. And finally, it delivers its first on-chip AI accelerator, the neural processing unit, also called the NPU. It's fantastic for offloading those long-running AI tasks that really allow us to improve battery life. The energy efficiency of the NPU is 2.5x better at executing the same code from our previous generation of products. 2.5x better, generation to generation. Intel Core is our most advanced integrated platform to date.
We will deliver the best AI PC experience. We promised you a leadership product as part of our annual cadence of roadmaps, and this new processor family delivers on all of it. As you can imagine, putting all of these new technologies into one processor was absolutely a heavy, heavy lift, but Intel Core Ultra represents our largest shift in architecture in more than 40 years. It was a radical change in how we manufacture and assemble our microchips, but it was absolutely worth it. Amazing technologies like Intel 4 , 3D performance hybrid packaging, and new microarchitectures. Intel Core Ultra beats competing processors by double digits. Delivering great performance for Intel, that's pretty straightforward, but we've heard from a lot of you, getting great power efficiency is a lot harder, but very important.
So the same architecture and technologies that allowed us to drive this amazing leadership performance also allowed us to take us to new heights on energy efficiency. And why does that matter? Versus our prior generation, we're up to 30% lower on processor power in things like web browsing, watching movies, video conferencing, things that we're doing all day long. And we even reduced Netflix by 25%, so you can browse and binge all your favorite shows from anywhere in the world.
That was Intel generation on generation. Now, let's compare to the competition. So when compared to the competition, we deliver up to 79% lower power in key metrics. It does not get better than Intel Core Ultra for an ultra-thin PC. It's the most efficient x86 product in this segment. And because we took the time and invested in this incredible efficiency, we were able to expand the size of the GPU for built-in performance to deliver the next level of graphics performance. Intel Arc Graphics brings all the modern technologies and experience from our discrete graphics into the latest processor. Not just gaming, media, everything. So with the Intel Arc GPU, we're seeing gaming and graphics performance that's up to 2x our previous generation.
More importantly, it's the hardware and the software innovation combined that enables us to deliver this built-in GPU with world-class performance in this space. But don't just take my word for it. I know you guys love to see things live, so let's welcome Jim Johnson to the stage. Welcome, Jim.
Appreciate it.
Show them what you got.
Let's show off a few of the capabilities you just mentioned. We chose a gaming workload because it puts the highest demand on the CPU complex, the integrated GPU, and the memory, memory subsystem. A company out of Poland, One More Level just released a title, Ghostrunner 2, getting rave reviews. On the left, you see it running on our prior generation product, and on the right, you see it running on Core Ultra. Just by moving to Core Ultra, you speed up the performance 2x, and by using the XeSS AI upscaling feature, you increase 3x. This processor runs at the same power envelope as its prior generation and delivers on the promise of power efficiency. It's one of 80 titles using this capability and this architecture and you can play it right next door in our demo showcase. Back to you, Michelle.
Awesome. Thanks, Jim. Just to recap that, we tripled our energy efficiency, and that's absolutely incredible, and something that we've been hearing from all of our customers that they absolutely want. Now, you're here to hear about the hero of the hour, AI. With Core Ultra, we're introducing the all-new NPU to enable low-power AI across the client. But I think it's important that in addition to the NPU, we architected our CPU and our GPU cores to deliver the most AI-capable and power-efficient client processor in Intel's history. Every engine on our chip can handle AI workloads. So whether you're using AI to edit your videos or create music, you're building images from scratch, all of these are going to be easier and much less time intensive.
So, for example, if you're a video editor, you're in Adobe Premiere, we've cut the time for you to edit a video by more than 56% versus previous generations. We know how many creators are out there. This is an absolute game changer for them. And with Core Ultra, you can do generative AI locally, so you can use something like Stable Diffusion and create a picture or a graphic that you want in half the time it used to take you. And we're very excited and look forward to Intel Core Ultra AI PCs running the exciting Microsoft experiences, like Copilot and Windows Studio Effects, coming to the market very soon. But it's not enough to have AI hardware. The real difference maker is going to be the partnership and the way that you bring software to your hardware platform.
We tested nearly 20 local AI models and workloads before launching this processor. The workloads spanned from video to image editing, Stable Diffusion, machine vision, teleconferencing, and so much more, all the things that you would expect. The Intel Core Ultra is the only processor that successfully ran every single one of these AI workloads. For users, that means compatibility that translates into frictionless and easy experiences with AI software on the PC. I've talked to many of you, and you all say, "What is the experience and the software that's going to come?" And we can show you right here, it already runs on Intel Core Ultra. And it's all because of this massive industry investment that we're making in the AI ecosystem. We recently launched the AI PC Acceleration Program.
This is a first of its kind, global initiative, really designed to increase and enable AI software development within the PC ecosystem, bringing those developers back to the heart of the PC. And this is a multi-generation program, where we're providing hardware and software engineering to really create and dream that next great AI-powered application. We don't know what that is, but I guarantee you, developers will absolutely figure it out. And the momentum to date has absolutely been incredible. As I mentioned, we're committed to delivering over 100 million PCs. That's the scale that Pat talked about. That's the scale that gets developers to come to your platform, and we expect to have 300 AI accelerated features enabled on the PC, and we've already started doing that work.
We're gonna have 100 software partners, partners like Adobe, Dolby, and Zoom on our platform, and we're on track to hit this milestone in less than a year. Many of them are available today, and if you go to the showcase, you'll see just a few of them. That's why developers are coming to Intel. We're frictionless. We support more models and more frameworks, giving developers a wealth of choice and a variety of tools and AI models that they can choose from. We don't tell them how to do it. We offer them the tools, and they can choose. Why would customers pick an Intel AI-powered PC? Because no one else in the industry has worked with more partners to bring features and experience to end customers. An AI PC, it translates to experiences, and those experiences are customized and specialized for each person.
And I believe that AI will scale the fastest because we bring it to the PC. We bring it to hundreds of millions. And while others love to talk about what's to come, we're delivering this impact starting today. Yes, exactly. You can buy it today. And let me just show you what I mean. We're gonna do another demo of a music production, and so I have the amazing independent music artist, composer, and performer, Kevehi, with me here. Hello, Kevehi. Nice to see you.
Thank you.
All right, so Kevehi, maybe tell people in the audience what you do.
So I am a one-woman band.
Well, now we're a two-woman band.
Two-woman band. I'm excited for it. So typically, I generate all of the tracks on my own. But recently, your engineering teams turned me on to the work that you're doing with Audacity. So now I can AI-generate drums, bass lines, melodies. It really cuts the time down by a lot.
Were you using AI to create your music before?
I was, but I just ... I didn't really have anything that creatively I could use until now, really.
Great. So what are you creating here?
Usually it takes me about a week or so to complete a project, but with the power of AI, I was able to do this song in under 24 hours.
Yeah, we didn't give her much time.
Yeah. 24 hours, right here in New York, I was able to just kind of create this track with the power of AI. So let me show you.
Yeah, please.
Using Audacity, I'm going to generate a very, very funky piano, and in just a bit, we're gonna hear what it comes up with. This is basically what you do. You just keep generating these new tracks, and you find inspiration somewhere. Okay, so let's hear what we got.
That is really funky.
It is funky. I feel like a parade-
Yeah
... bringing wear it now. Yes, it's really great-
How long would it have taken you to create and iterate on these tracks before to kind of get to what you wanted for your piece?
I mean, oh, at least a week or so to complete just one project now.
So we gave you about six hours of time, and I know you say you're going to gigs all the time, so being able to do it locally and on your PC really is important to you, where you may not have great network or be able to connect to the cloud and get that boost of performance.
Absolutely. It's portable creativity powered by-
Awesome
... AI and Intel.
All right, so did you create a piece for us?
I did. Do you guys wanna hear it?
Yeah!
All right, let's do it. I'm excited. Oh, sorry!
That's okay.
Let's go in the beginning. There we go. Now it's the front.
Okay, are you guys waking up now? That's awesome. All created in less than really 12 hours.
Yeah, really.
'Cause I'm assuming you slept last night.
Yes, just a little bit. After I had some pizza, I had a good night rest.
Awesome. This is fantastic, and it's just one example of how all these creative minds can use AI to really hone their craft, and really change it, and bring it to life.
Exactly.
With that, Kaveh,
Oh, my-
I want to give you the first
Intel Core Ultra PC.
You have AI in your hands starting today-
Thank you
... to go and create and make amazing music.
I better get busy.
Thank you.
Absolutely.
Thanks so much.
That's just one example of all the time-saving goodness that's available starting today. We have a full range of Intel Core Ultra processors, and you can expect to see more than 230 designs from our global OEMs, including Acer, ASUS, Dell, HP, Lenovo, LG, MSI, Samsung, and more, and many are on the show floor for you to take a look at. As well as, in addition to starting today, you'll be able to buy these machines in select retail stores, including Intel's latest Intel Evo Edition laptops, and you can go just down the street in New York City to your local B&H retail store and get one today. So if you have a Christmas list that you need to fill, that's where you do it. Yeah, exactly. They're also gonna be available online at select retailers: Amazon, Best Buy, Costco, Newegg, and Walmart.
All of your favorites, you can buy them starting today. The momentum, it doesn't stop here. You're gonna hear more about Intel-powered vPro commercial systems with AI coming to market in Q1 2024. Our supply chain is strong. We're firing on all of our cylinders, and the excitement is growing across the industry for these new Intel-powered AI PCs. Don't just take my word for it. Look at what our customers are saying about this product. They told you the best is yet to come in the earlier video, and it arrived today. I'm so excited, if you can't tell. Yes, we hit on some important themes today, and this is such a momentous occasion, with hundreds of thousands of hours of work coming together across the amazing Intel global teams to bring you Intel Core Ultra. So let me do a quick recap.
This is the largest architecture change in the last 40 years, the best AI experience for PCs, a world-class Arc GPU, a dramatic improvement in power. Intel Core Ultra is the best for ultra-thin laptops. Just as Intel is transforming the client market with powerful AI capabilities, we're also transforming the data center market. Please welcome my friend and colleague, Sandra Rivera.
Thank you so much, Michelle. So moving from all the exciting news around the AI PC, I'm gonna talk about everything we've been doing in the data center and server business. So at Intel, we've been on a journey to evolve our data center platforms to align with customers' needs today, so they can plan for the future. Over multiple generations, our investments have focused on delivering hardware and software innovations tailored to provide leadership in the workloads that our customers care most about, and today, that workload is AI. Intel started building AI acceleration into our Xeon processors more than a decade ago. Today, Xeon has established itself as the industry standard for powering the AI workflow. In 4th Gen Xeon, we integrated a revolutionary technology called Advanced Matrix Extensions, which boosts AI performance by up to 10x .
Most of the world's leading companies have adopted 4th Gen Xeon to accelerate their AI, including all of the top 10 hyperscalers. Of course, we're seeing adoption beyond CSPs as well. The South Korean search company, Naver, for example, adopted Xeon as a cost-effective alternative to GPUs. The demand for AI solutions has never been greater, and Intel remains firmly committed to bringing AI everywhere by delivering the hardware and software that makes it easier to deploy. Today, I'm pleased to launch our next generation 5th Gen Xeon processors. This is the most power efficient, highest performant, and security-enabled Xeon we've ever delivered, offering a 21% average performance gain over the previous generation. Privacy and control over data is paramount, and 5th Gen Xeon offers increased confidentiality and security with Intel Trust Domain Extensions, which will be generally available to all OEM and CSP solution providers.
AI offered all within the same power envelope as the previous generation. 5th Gen Xeon is the best data center CPU for AI, period. We have built AI accelerators into every one of the processor's 64 cores, giving users up to 42% higher inference performance versus the prior gen. This means enterprises can easily and cost-effectively use a CPU to run the latest generative AI models, like GPT-J, DALL-E, and Llama 2. 5th Gen Xeon can fine-tune large language models like the 1.5 billion-parameter BioGPT model in just minutes, while meeting critical SLAs for the 70 billion-parameter model, Llama 2, when scaling out to multiple nodes. 5th Gen Xeon is deployed on the same platform as the previous generation, and purpose-built to scale and bring AI everywhere. In the cloud, AI-powered natural language processing applications will operate up to 23% faster.
In the network, 5G user plane functions will flow an average of 50% faster. At the edge, objects can be classified up to 24% faster. These savings enable businesses to operate smarter and at lower costs. For years, we've worked closely with the ecosystem to accelerate the adoption of AI in the enterprise. Our collaboration with IBM is a great example of this work. With over a decade of experience from its Watson solutions, IBM has evolved its AI to deliver its Watsonx platform. Watsonx enables businesses to customize and scale AI based on their unique data. I've invited Vikram Murali to talk about their use of Intel's AI solutions to help power Watsonx. Let's take a look.
With Watsonx, you can leverage foundation models for generative AI and machine learning using your data wherever it resides, to advance responsibility, transparency, and explainability. watsonx.data is IBM's fit-for-purpose data store built on an open data lakehouse architecture, and it is truly hybrid solution available on both private and public clouds. We have been working closely with Intel on benchmarking improvements to query performance, and are excited to announce fifth generation Intel Xeon processors have up to 2.7x better query throughput on watsonx.data than compared to previous generations during internal testing. We are looking forward to continuing our long history of collaboration with Intel to optimize data and AI products for our customers.
Thank you, Vikram. I'm excited to see how enterprises will use watsonx to accelerate their business outcomes. The features we've built into 5th Gen Xeon enable customers to lower operating costs and advance their sustainability goals. Customers upgrading legacy Xeon servers will experience up to a 77% TCO improvement with AI workloads. We work closely with our OEM partners to put these innovations into customers' hands. Servers based on 5th Gen Xeon will be broadly available starting in Q1 of 2024 from leading providers, including Cisco, Dell, HPE, IEIT Systems, Lenovo, and Supermicro. As more enterprises move to cloud-based environments, our collaboration with the world's hyperscalers is a strategic advantage for many businesses who want to access Xeon's unique features in the cloud.
For over a decade, we've worked with Google Cloud to ensure our mutual customers have early access to the latest services offered by our data center solutions. Now, let's hear from Mark Lohmeyer to talk about our work together.
At Google Cloud, we deliver workload-optimized infrastructure that enables our customers to harness the potential of AI-enabled applications, cloud-native workloads, and demanding enterprise workloads to achieve their business goals. Google Cloud and Intel share a rich history, dating back to our introduction of Intel Xeon first gen processors in Google Compute Engine, which was first to market amongst all the major hyperscalers. Since then, we've achieved numerous groundbreaking firsts together. Earlier this year, we introduced the general availability of our Google Compute Engine C3 instances, powered by the fourth gen Intel Xeon processors. This was the first of its kind amongst the top hyperscalers. C3 also supports Intel's AMX technology, delivering exceptional performance, security, and cost-effectiveness. By adopting Google Cloud's C3 virtual machines with the new AMX instruction set for AI, Palo Alto Networks experienced a remarkable 2x performance boost in its inline models.
We're even more excited about what's coming next: new Google Cloud virtual machines powered by 5th Gen Xeon processors, enabling our customers to harness the latest performance, efficiency, and security benefits.
Thank you, Mark. Thank you, Mark and team Google Cloud. We're looking forward to what you have coming with Fifth Gen Xeon early next year. So across on-prem, hybrid, and cloud environments, companies are experiencing firsthand the leadership TCO that Xeon delivers. One such company is Gallium Studios, a leading indie game studio. The team at Gallium were looking for ways to meet the high throughput, low latency demands of real-time natural language processing applications. And now, I'd like to welcome to the stage the CEO and co-founder of Gallium Studios, Lauren Elliott. Please join me. So welcome, Lawrence. Thank you for being here with us today.
Thank you.
So can you talk a bit about how Gallium Studios and is using AI in your, in your workflow, in your applications?
Sure. Gallium Studios is an independent gaming company that I co-founded with Will Wright, who is, the author of a lot of games, Sims and SimCities amongst them. And inspired how memories are stored in your brain, we're building a new AI simulation game called Proxi, and where players, enter their memories as they're visualizing them as 3D globes, and these 3D snow globes are full of imagery and animation and everything that makes the memories. From those memories, we create the Proxi personality, and intelligence, and we bring that to life so that the players can interact with them. We like to say you can learn from it, and it can learn from you. This is actually a simulation running live of memories.
It's our simulation, and those memories are all connected to one another, as the brain does, and so it's running live, and that's kinda cool.
So all these simulations must use a lot of AI compute power.
Right.
How do you talk about just the way that AI is helping with this process?
Yeah, as you can imagine, the whole process is full of a bunch of AI and large language models. We began building Proxy, we were using external APIs into these services, but they were slow and costly, and they didn't scale for us as a gaming company.
Mm.
So we decided to bring everything internal on our own GPUs. That did improve the speed and latency, but the cost of the GPU-based cloud instances was way too high for a gaming company.
Why did you choose Numenta?
Ah, good question. So we moved to the Numenta platform for something they call their intelligent computing. The platform is designed specifically to deploy and scale LLMs on GPUs and not CPUs, and it gets the best performance on Intel chips. e started working from Proxy, at Proxy, as a kind of a neuroscience-
gaming company. Numenta's approach with their AI research-based orientation resonated strongly with us. So the match was perfect, and we're excited to work with them on their platform to bring Proxy to life.
So tell us what kind of results you got?
Well, that's the great part of it. Once we switched to Numenta and the Xeon processors, we're able to run inference about 6x - 6.5x faster than their previous... than our previous GPU solutions. And that brought the cost down significantly, so, like, we could bring it on mobile devices, and also reduced the latency of creating these thousands of memories that we have to do in real time for the players.
6.5x faster.
Yep.
That's pretty amazing results and pretty amazing savings. Yes. Well, I can't wait to experience Proxy firsthand. Thank you so much for being with us today, Wright.
Thanks. Thank you.
Well, we know that when it comes to AI, there is no one size fits all. AI is a complex workload and requires heterogeneous platforms to meet customer requirements. Today, Xeon is the industry standard CPU for powering AI across the entire workflow. But for training and deploying large-scale generative AI models, we have our Gaudi 2 accelerator. We've demonstrated Gaudi 2 delivers leadership price performance compared to the most popular GPUs. Based on the most recent MLPerf training benchmarks and estimated range of server pricing, Gaudi 2 delivers between 40%-50% better price performance than NVIDIA's H100 for training the 175 billion parameter GPT-3 model. That's pretty good. For decades, Intel has been on the forefront of extreme scale and performance, and we're bringing that experience to AI. Using Gaudi 2 and Xeon, we're building one of the world's highest-performing AI supercomputers.
This system will leverage many of the same design and system principles behind the Aurora supercomputer, one of the world's most powerful AI and HPC systems in the world. The key to scaling AI and bringing the technology everywhere requires a platform approach built around a common software stack of and foundational silicon, and a pervasive network based on open standards. This allows for greater programmability, model portability, and ecosystem adoption. We've optimized our AI platform at every level, from silicon systems through all layers of the software stack, all the way up to the application layer. Our approach allows developers to use the tools and language they're already familiar with and are using. PyTorch, TensorFlow, DeepSpeed, Hugging Face, and many other industry-standard frameworks and libraries have Intel optimizations included as default.
As the lines between cloud and edge blur, we see the increased deployment of smaller, more nimble AI models that span cloud, edge, and client environments, as you heard from Michelle. Hybrid AI is rapidly becoming the most effective approach to deploying AI everywhere. Hybrid AI lets users deploy AI models wherever it makes the most sense, based on latency, cost, or regulatory considerations. With Intel's software tools and foundational hardware, developers can build unified code that targets CPUs, GPUs, and AI accelerators, and they can scale their AI models from the cloud to the network to the intelligent edge. We also work closely with leading software providers to validate and optimize their software on our platforms. This work helps to further accelerate the deployment of highly optimized AI solutions.
Intel and VMware recently announced a collaboration where we're delivering a jointly validated AI stack, and this solution allows customers to use their existing infrastructure and Intel software suite to rapidly build, deploy, and secure their AI models. Through our work with this broader ecosystem and AI community, we're delivering state-of-the-art solutions to train, fine-tune, and deploy AI models on Intel silicon. Let's hear from a few of those community members now.
I think one of the biggest opportunity and challenge that we have right now is the ability for AI to translate into business value, and that is what we try to unlock as part of our solution. Intel CPUs are the ones that our platform and solutions are optimized to run on, and we also take advantage of tremendous capabilities and optimization that have been performed on the Intel Core hardware platform itself, for us to run our AI workloads. So this has really helped our customers, in terms of realizing value quickly and efficiently.
As a platform, we allow data scientists, MLOps, and IT to actually use their infrastructure to run their AI workloads, and we leverage a lot of Intel technologies. Most of the processing by our customers is happening on CPUs, but there's also a wealth of solutions from Intel AMX, the oneAPI solutions, or Habana Gaudi, that can be accessed via the platform by our customers. So they allow us to actually provide more processing for these specific AI workloads, offering better performance to our customers, which translates in them getting their models faster in production and creating business value faster.
Hugging Face is the number one open platform for AI builders, with over a million open models, data sets, and AI applications. We've partnered with Intel for many years to democratize machine learning as efficiently as possible, to accelerate AI from the hardware to the software layer through open source collaboration. Intel Xeon CPUs are the workhorse for inference of efficient models, and together with Intel, we've enabled great new generative AI use cases.
At Roboflow, we believe anything that can be seen will be turned into software with AI. Over 400,000 developers, including those from half the Fortune 100, build with Roboflow's tools to enable computer vision to be in their products and services. That includes companies like Walmart, Rivian, and Cardinal Health. Intel is a key component in how we enable our customers to deploy production AI. From Xeon, to OpenVINO, to Habana Gaudi, Intel enables our customers to deploy production AI workloads, both on the edge, in real time, and on the cloud.
That's just a small example of all the work we're doing with the software ecosystem. We've covered a lot of ground today, but that's because we had so much to celebrate with the launch of 5th Gen Xeon. Let me summarize. 2023 was an incredible year. We kicked it off with the launch of 4th Gen Xeon in January, and now we capstone it with 5th Gen. We're proud of what 5th Gen Xeon delivers. It is simply the best CPU for AI, period. 5th Gen also arrives with significant leaps in performance and efficiency, all within the same platform as the previous generation. 5th Gen offers greater security with enhanced privacy and control of your data, with the broad availability of Intel TDX. And beyond the hardware, Intel is committed to continue delivering high performance, ready-to-use, open software ecosystems and environments.
Intel is truly bringing AI everywhere. Working together with our partner ecosystem and software providers, we're scaling AI in the data center, network, and across a broad range of vertical edge markets. Now, here to share more about our partners and how customers can get their hands on these solutions is Christoph Schell. Welcome, Christoph.
Thanks so much. All right, so I'm going to bring this home, and you heard from Sandra and from MJ about the great products and the systems that we're building. I'm here to tell you how we are thinking about this from a go-to-market point of view. For the PC, for the edge, for the networking, for the cloud, and for the data center, it takes a village to bring AI everywhere. And I'm going to use one of these to double click, and that's the edge. And the reason why I'm choosing the edge is because it is the fastest growing segment for compute. And if you look at my funnel, this is where a lot of action happens. This is where a lot of customers are engaging with us.
The reason why they are engaging with us is because they can drive productivity on and really competitive differentiation at the edge. It's very important to think about this in vertical industries. I'll give you a couple of examples. Healthcare. Think about what ultrasound can deliver versus what the human eye can see. Think about an energy management system, a smart grid, how to effectively transport energy from point A to point B with less redundancy. I'm from Stuttgart, Germany. You can hear my accent. For me, life is all about automotive. Think about EV-powered cars, self-driving cars, and how these cars become edge devices, how these cars can be connected. For my team and I, super important to have vertical expertise. The edge represents $445 billion of global market opportunity. This is huge!
Within the edge, AI is the single fastest growing workload. Intel has been investing in the edge since 2018. You saw one of the products, OpenVINO. That was the first product that we really took as a platform, and we have been investing. And what is cool about this is that we have so many developers, system integrators, ISVs, that are innovating together with us, and it takes that village of an ecosystem to make that happen. Because of this investment, we have 90,000 edge deployments to date, 90,000 . The wealth of experience that we have from these projects informs how we are deploying AI at the edge now. And if you are a developer, the one scarcity you have is talent. So you have to think about, "Where do I spend my time?" And Intel is this one-stop shop for a developer.
You can innovate on the same software stack, as Pat said, whether that's Intel Core Ultra on the PC or the fifth gen Intel Xeon for net, and networking and edge. It's one-stop shop for everything that you can build on. Today's innovation is an absolute breakthrough. And what I want you to understand is, this ecosystem is important, but it's also really second to none on how we engage with customers. And I have one video of one customer that we've been working closely with for the past five years. It's a company that brings video conferencing to over 300 million users every day. The company is Zoom, and here with me today on video is Zoom's Chief Technology Officer, Xuedong Huang. Let's hear from him.
AI is an important technology investment for Zoom, which is why it's been a part of our offering from early on. Our customers have benefited from Zoom AI through high-quality noise suppression, video optimization, virtual backgrounds, and live translation. Intel has been a strategic partner for AI across our client and cloud offerings for years. It's critical to our business, since Intel products help us manage the cost and the complexity by selecting the right product for our AI workloads. By using Intel software, like Intel Extension for PyTorch, we get optimal performance out of Intel Xeon servers. It helps us to stay flexible and enable Zoom to continuously improve the effectiveness of our AI models. Zoom is proud to partner with Intel to help make human-to-human interaction more meaningful, streamline business processes, and improve the impact of communication with coworkers and customers.
With running AI workloads on client or in the cloud with Intel, we know our customers will have a great experience.
Thank you. So Zoom is a really good example to illustrate how you can engage with Intel across the entire compute spectrum, and we have been doing this for five years. AI has been at the center of our collaboration, and we will continue to work with Zoom. Now, I come back to this ecosystem because it's such a huge advantage that Intel has. The breadth of that ecosystem, the global reach of that ecosystem, are second to none. It's very robust, and it's based on very open and transparent standards, and that's what the ecosystem partners like. This ecosystem was also very foundational to how we are engaging with another customer, a company called Samsung Medicine. They are a pioneer in healthcare innovation. They have ultrasound imaging applications that use AI for really driving effectiveness for patient care.
Previously, when we worked with them, we engaged on an Intel Core CPU and a competitor's GPU, so two different products, two different companies. When we started talking with them about Intel Core Ultra, they first didn't really want to engage because it had a built-in GPU, and then they saw the performance. They saw the performance of an integrated product. Think about a new generation CPU and a discrete GPU in one combo. What they saw blew them away. They were able to really bring AI features to the ultrasound devices on one CPU without the need for CPU. Their cost reduced significantly. And just think about what this means for our customers. It's an absolute game changer in how they think about architecture, how they think about compute power going forward.
It has a huge impact on how we think about the addressable market that we will have in the future. AI everywhere, from client to the edge, from networking to the cloud. Today is a huge day to make a major step towards this vision. Why do Intel customers like us? Four reasons, and I was an Intel customer for 25 years before I joined Intel, so I actually know what I'm talking about. They like our open approach. They like the fact that we embrace open standards, that data is accessible. It's not in a black box. They know what happens to their data. They can innovate with their code, with their innovation on top of our stack. They really like the depth of our engineering. They like the fact that they can talk to our hardware engineers and to our software engineers. They like this ecosystem.
It's a go-to-market for them, ready globally, and they like our execution. They like that we're predictable. So 2023 has been a huge year, and I'm so much looking forward to what is going to happen in 2024, and I really invite all of you to go out there, go shopping-
Hey, hey, hey!
Pat, you're not supposed to be on stage right now. This is my thing.
I'm the boss.
You're the boss?
I get to-
Well, then come back on. What do you got?
Well, I'd like to just have one other little thing to show off here, and they just brought it out of the lab for me.
I saw that, yeah.
The first ever Gaudi 3 in public, so here we go.
Wow!
The next-generation accelerator. So Gaudi 3, out of fab, in the lab, being powered on, looking healthy, and not just Core Ultra, not just Gen 5 Xeon, but the next generation of accelerators. We are executing like crazy. This is great, Christoph.
So, team, and you're obviously all watching, our quota has just gone up for next year, okay? But I'm very happy for our engineers to deliver this. Thank you so much.
Thank you, Christoph.
Let me try and bring this home now.
Okay.
Super excited about 2023. You saw that our PCs and our fifth gen, they are not just on PowerPoints, they are real, okay? So please go out there today, do the Christmas shopping, get your holiday gifts, okay? And why don't you close us out?
Hey, you know, and AI everywhere, right? And Core Ultra, and Gen 5 Xeon, and the next generation of accelerator. Working with our partners, as you saw today, the ISVs, the ecosystem, the excitement that this is bringing, truly making it part of every one of our lives, truly changing humanity. Thank you so much for joining us today.