Welcome to the heart of Silicon Valley, Intel Innovation 2022. The doors are open, and thousands of developers are filing in to see the opening keynote with Intel CEO Pat Gelsinger. After days of behind-the-scenes work, the stage is set and everything is ready for a two-day event where tech geeks gather, collaborate, and share their plans for the future. Innovation 2022 promises to be a groundbreaking event, and it all starts now. Welcome to Innovation Live 2022, coming to you from the San José McEnery Convention Center. Hi, everyone. I'm Madeline Littrell, your host for the next two days, broadcasting live from the Innovation Zone. It is wonderful to be back in person with developers from around the world who are here to experience all the hands-on demos, workshops, technical highlights, and of course, the keynote and luminary presentations. It's going to be an amazing two days.
I'm not alone this year. The show has gotten so big, I have a partner in crime who is joining us now from the Intel blue carpet backstage. The second half of our broadcasting team, my new BFF, Ravi Dosanjh. Welcome, Ravi.
Wonderful to be here with you, Madeline, our attendees, and the folks at home joining this year's Intel Innovation event. I'm standing out here on Intel's version of the red carpet, the blue carpet, and excited to geek out with our execs, developer community, ecosystem partners, and of course, you, over the next two days.
Looking great, Ravi. What's the energy like? Who are you seeing back there? It's exciting.
Well, I tell you, I've had the chance to speak to many folks. Some folks at the media lounge, TechInsights ballroom, the Innovation Zone, and everybody here backstage, they're all excited to be back in person this year and eager to see the keynotes and technology demos.
Thank you, Ravi. I'm with them. I'm eager, too. We'll be checking in with you throughout the show to see who you're rubbing elbows with. Everyone at home, stick with us. We'll be bringing you some exclusive sneak peeks and interviews ahead of the live keynote, and you won't get this anywhere else. All right, let's take a look at today's full schedule and all we have in store for you. First up, you'll get an exclusive first look around the showcase floor. We'll meet the winners of Intel's AI Global Impact Festival and get to know Ria Cheruvu, Intel's own pioneer of AI ethics. After the keynote, we're excited to welcome Pat Gelsinger for a live debrief. You won't want to miss that one. Also on tap are the must-see highlights from the keynote and an interview with Intel's Chief Technology Officer, Greg Lavender. Also, tweet at us.
Use the hashtag #IntelON to join the discussion and for a chance to be featured on the show. For the first time ever, we are so excited to bring you backstage access and exclusive content for our virtual audience. That's you. This is live TV, so be ready for anything. Speaking of, I'm hearing that Ravi found a very special guest who stopped by on his way to the main stage. Ravi?
Thanks, Madeline. You never know what's going to happen backstage, and you never know who I'm going to intercept on their way to that stage. Pat Gelsinger, Intel's Chief Executive Officer. It's great to connect with you again here at Intel Innovation 2022.
Hey, Ravi. Always a pleasure to be with you, and I appreciate your enthusiasm, excitement, and we are getting ready to kick off Innovation 2022.
We sure are. I can feel the buzz. Can you feel the buzz, Pat?
Oh, absolutely. The demos, the show floor, all of our vendors and partners showing up. It's going to be just a fabulous couple of days together.
Amazing. Now, Pat, you went viral last year for doing something really unique as a pre-show get ready. Are you doing anything this year that we should be thinking about?
Well, like always, hey, I get up early, I practice my keynote, I get my speaker notes all ready, and that's sort of like religion for me. Then I don't look at them again until I go on stage, right? That's all done. Of course, you know, hey, you need a few push-ups with me, Ravi?
You know, I don't think I'm going to join you.
Okay. Can I do a few?
I think you can.
Let's do it. All right, great.
Wow. I think we're at 11.
Okay. That's, I'll do some more later, I promise.
Nice job. I've got to talk to you about that form, but hey, that was a torrid pace. Well, Pat, we love your energy. Your keynote's coming up in 30 minutes. Anything that the audience should take away or any sneak peek you want to give us about your keynote?
Well, maybe, three things. You know, one is the new product introductions, right? Obviously it's a geek fest, but you know, coming out with cool new technologies and being able to describe them to the audience. You know, second, developer. Let me remind you, developer. The heart of what we do is enable hardware and software developers. You know, that's what I'm going to talk about, Greg's going to talk about. You know, and between. You know, I think of it as a keynote in two parts, right? I have some things and Greg has some, you know, aspects, but this idea of open. We are here for the open ecosystem, the choice that we enable across the platform and then building more trust. That's number two. Number three is, hey, some of the fun guests I have.
You know, I am such an underachiever. I'm going to describe that on stage. One of my lifelong legends is joining us here today. You know, we have some really great tech, some really great messages to the developer community, and some really surprising people joining us here at Innovation.
Excellent, excellent. I don't know about that underachiever bit, but I got to tell you, Pat, you've been on a torrid pace since last year's Innovation. That was not in person. It was virtual. Now we're here in person. Tell us how you're feeling about the energy of everybody being here in person and yourself being able to engage the developer community.
Well y ou know, it's been just a great, you know, six to eight weeks. You know, obviously the White House, the bill signing on the CHIPS Act, Ohio groundbreaking, where we, you know, got it all underway, and now being together here in the Silicon Valley, you know, where it all began and what Intel really helped to usher into the world. All of this to me, you know, today is such a great day to be back together in person with the developer community in Silicon Valley.
Excellent. Well, Pat, good luck in that keynote. We'll see you throughout the show.
Thank you, Ravi.
Thank you very much. Madeline, what's shaking back at the host desk?
Ravi, I can't believe you wouldn't do pushups with Pat and that you called him out on his form. Bold move. It does seem like Pat is ready to go, so thank you. We'll check back in to see how you're doing in a few minutes. Okay. For everyone watching online, we don't want you to miss a thing. Ravi and I are going to do our best to be the next best thing to being here in person. Consider us your eyes and ears and your tour guide around all the action. Yesterday, I checked out some of the demos here in the Creator Zone and Showcase, and you'll be amazed at what I discovered. Hi, everyone. You can see that we're still setting up here at our Innovation Showcase, and I am reunited with my friend and demo nerd, Craig.
We're so excited to have you back with us.
Good to see you, Madeline. This is great.
Always a pleasure.
Exciting time at Innovation.
Amen to that, sir. Okay, tell us about the Spark Theater. This is like a mini keynote area.
That's absolutely. We're bringing this down directly to the Innovation Showcase, and here we are. You'll see all the demonstrations that we have, over 100 demos on the floor.
Mm.
At the center of it is our beautiful Spark Theater here, which we're bringing in a ton of mini keynote talks from everything from demonstrations to executive speakers, to some of our most valued partners, as well as customers are coming in to go ahead and tell us all about the brand-new innovation that they're seeing with Intel technology this year.
Love it.
Stop on by. You've got to check it out.
I love it. This was the first stop, now on to the next one?
Let's go.
You have some time?
Absolutely.
All right.
Let's go check it out.
Let's do it. Craig, tell me about the Dev Toolshed behind us.
The Dev Toolshed. This is something here at Innovation where we have Intel experts to work through all of our open source tools, VTune, OpenVINO, all of the tools that you have to really maximize the performance of your code on Intel hardware. We also have an entire hands-on expert set where Intel employees with all of their expertise will be able to help you work through your code, your projects, to be able to make sure that they're optimized in the best way. Pretty cool.
We were saying earlier, this is like BYO problem.
Bring your own-
Like BYOP.
Bring your own problems. We love to solve it. Developers, come on down to Innovation. We've got you. All these new tools, let us know how to use them.
Love it. We love that confident stand.
That's pretty cool.
Over here in the Innovation Showcase. Well, you have to get to a rehearsal.
I have a keynote. Oh, man. Gosh, I'm sorry.
We'll see you in a few.
Great, Madeline. Thanks, everybody.
Bye. Thanks for stopping by. Well, we'll see more of Craig in just a few minutes during the opening keynote with Pat. In the meantime, we have a few more stops. Let's head to the next one. Oh, that's such a good dog. Shh, I think she's sleeping. We'll come back in a minute. Hi, everyone. I'm here at the Variable Viewpoint Capture with Horst. Tell me, what's going on in this booth?
Behind us, we have a setup of a multi-camera system that we developed. It is a consumer-grade, 15-camera system that captures 15 different views of what you're looking at in front of you. Then we have an entire end-to-end infrastructure that takes that content, streams it to the cloud, and turns it into immersive video that it can view at different locations here in the demo.
It's incredible because you don't need special equipment for this. It just runs on a normal laptop, and it responds to your face and your movement, even spatial audio, right?
That's right, yeah. I mean, you use a standard laptop. It tracks your head location, knows where you are relative to the screen, and then it creates a view that's exactly the view that you need from that perspective. You're using the same set of cameras. It's still only 15 cameras.
Mm-hmm.
You're creating 48 different views that are then shown on a special TV that creates 45 views simultaneously. It does look like a hologram. It does look like a floating face in front of you.
Well, thank you so much for being here with us today. This was a really neat demo.
Thank you.
On to the next one. You're awake. Come here. Come on. Oh. Oh, you're so good. I think this one's probably my favorite of the whole showcase. I know I said it on the package, but I think that robot dog might be my favorite technology in this space.
Yeah, he might pay a little bit more attention than my Skippy at home. Yeah, honestly speaking, Madeline, this is mind-blowing stuff, right? I got to tell you, it's great to be back here in San José, California, with the company that put the silicon in Silicon Valley.
That's right. There are so many experts and luminaries and pillars of the developer community gathered here, and they're geeking out with each other, and we get to experience it in person live.
Yeah. Lots of stuff going on. Lots of deep engagements coming up with showcases, demos, brain dates, so much more to experience and see. Certainly more than can be packed in two days.
100%. We're packing a lot into the next two days, and we're so excited to see what happens in this keynote live, and then what happens again in day two with Greg Lavender. All right, Ravi, who do you have with you now?
Well, none other than Michelle Johnston Holthaus, Intel's Executive Vice President and General Manager of the Client Computing Group. Michelle, it's great to see you again.
It's great to be here, Ravi.
Yeah. Thank you. The energy is, like, so good back here backstage. Michelle, Intel is dropping a lot of big things this week, right? I wanted to ask you, would you be able to give us a little bit of a hint, a little bit of a sneak peek of what we're gonna be experiencing this week?
Well, what I can tell you is Pat has a lot of exciting things to announce today, and I'm not gonna drop any of the trade secrets, but I can tell you that there might be a few things with the PC and next leadership products that I think you're all gonna be very excited to hear about.
That's amazing. We had a chance to catch up yesterday, and I know you're not only excited to spend time with me, but our developer audience here. Tell me about what it feels like to be here in person, rubbing elbows with our developer community.
Yeah. It's so exciting just talking to them, the passion that they bring, and it's our job to make sure that we deliver the tools and technology so that they can bring our silicon to life. They are able to bring that silicon to life in ways that we never imagined. It's thrilling to be here this week back in person with some of the best and brightest.
What is some of the things you want the developer community to take away from the keynotes, the show, the engagements? Like, what are your top threes?
Yeah, when I think about the developers, I really want them to know that Intel is more than committed to an open ecosystem. I think that is so important. Where there's choice, where there's innovation, we know that wins. Intel is gonna continue to make sure that we deliver leadership products on a yearly cadence that they can innovate around. When we work together, we know we win with an open ecosystem.
That's amazing. When we talk about ecosystems, we talk about software, hardware. You did a thing with Stu and Zane on unifying the ecosystem supply chain. There's so many partners involved in this amazing ecosystem we have. Any last takeaways for you for them in Intel Innovation 2022?
It's gonna be an exciting week. We've got amazing products. You're gonna see those products come alive because of some of the developers in the audience, and I'm super excited to see what they create.
Excellent. Michelle, thank you for spending time with us. We'll see you a little later in the show.
Thank you.
Madeline, back to you.
Thanks, Ravi. Lots of emphasis on openness this morning. I think we're gonna hear a lot about that in the keynote. Okay, we're counting down the minutes to Pat's keynote. Coming up, we'll have more backstage interviews, and next, Intel's AI Global Impact Festival winners join us on set. What are you hoping to see in today's keynotes? Let us know by using the hashtag #IntelON. We'll be right back.
How many tacos did you order?
100.
The technology inside, the technology that moves the world. How wonderful is that? Intel.
Welcome back to Innovation Live. We're coming to you from the San José McEnery Convention Center, where everyone is gearing up for the opening keynote. For those who are just joining in, I'm Madeline Littrell, and I'll be your host for the next two days. As you might expect, we'll be spending a lot of time talking about AI in the next two days. A big part of that conversation is encouraging and celebrating the ones who will be shaping the future of artificial intelligence. Intel launched the AI Global Impact Festival as part of its commitment to Tech For Good. The festival brings together next-generation technologists, future developers, policymakers, and academics who work to solve real-world problems using AI. This year, Intel received more than 1,000 entries from 25 countries. Joining us now are the winners of this year's festival. We have Dina Marie, Krish, and Tanapat.
It's so nice to meet y'all. Thanks for being here with us today.
Thank you for having us.
Yeah, it's wonderful being here.
Awesome. Congratulations again. I'd love to hear a little bit more about your projects. Dina Marie, let's start with you. Ladies first.
Well, our project is indoor industrial safety and hazard prevention. We're really inspired because two of my team members have worked in industrial areas, and safety is just such a major concern.
Mm-hmm.
We really wanted to focus on how to use AI to prevent and predict any type of hazardous situations that might occur and to see, you know, who might be wearing their PPE...
Mm-hmm.
...versus who's not without the use of facial recognition.
Right.
We're using lidar for that.
Oh, very cool.
Yeah.
Using AI to protect people.
Yes.
Very cool. Krish, what about you? Tell us about your project.
Yeah. My project is basically an AI model which is aimed at improving the accessibility to programming-
Uh-huh.
by breaking the barrier of language to get started with programming.
Okay.
It basically summarizes the code which user writes in their own mother tongue, so they can better understand what they're writing. It helps people to get from non-native English background to get started with programming.
Oh, that's amazing.
Yeah.
How old are you?
I'm 17.
You're 17. I can tell you've done this before. You have great energy.
Yeah. Thanks.
He's like, "Yes, I love doing this." Awesome. Tanapat, tell me about your project.
My project is about I create a medical innovation that allows people to have a regular heart checkup because heart disease...
Wow.
...is very important in this world because it can happen to anyone and anywhere. We need to take tests about the heart, so when we have the regular heart checkup, we can screening the heart disease, and we know faster and get the treatment faster than normal.
You can do this outside of a doctor's office or a hospital?
Yeah, we can do it at home by ourselves.
Wow, that's incredible, especially for those who might not have direct access...
Yeah.
...to a hospital who might be in farther, more distant or suburban areas.
Yeah.
Very neat. Okay, what do y'all think are some of the biggest risks or challenges that you're going to face in this AI field?
I guess how responsibly we are using an AI would be a very big challenge in the future as well with our own projects. Letting the right use of AI with the right ethics is a...
Mm-hmm.
...big challenge ahead.
Yeah, responsible AI is going to be huge.
Responsibility.
What's next for y'all? Anyone wanna share a dream or what you're gonna work on next before we wrap?
In my dreams, in the future, I want my projects to using worldwide. I think I will expand to the rest of my country, Thailand, to-
Mm-hmm.
Allows everyone to get a regular heart checkup. Starting to expand to the rest of the world because, like I'm talking now is maybe someone die from the heart disease. Like, according to the CDC, every 34 seconds, maybe someone die in the United States from the heart disease, so it's very important.
Absolutely. Well, you are an inspiration, and no doubt you're gonna make that happen.
Yeah.
All right, y'all. Congratulations again. Thank you for being here today. Ravi, we're gonna take it back to you. Who do you have with you next?
Hey, Madeline. Well, I am standing here with Sandra Rivera, who leads our Intel Data Center and AI Group, and I got to tell you, I'm so excited. It's almost go time. Sandra, you're gonna have to hold me back.
I know. It's T minus 10 minutes, Ravi. Pretty exciting.
Well, hey, we're back here in person for the first time. How are you feeling about this year's in-person elbow bumping with the developer community?
Well, it is always better to be in person and to see everyone in 3D. Super excited. I actually got a little bit of a sneak peek in the Tech Showcase and Innovation Zone earlier today, so I can't wait to see everything that Pat's going to talk about, and then of course, to open up those developer podiums and platforms and demos to everyone.
Right. You got a little sneak peek. Anything you want to share with us?
Well, I can't really say too much other than for sure we are trying to make the innovations that we build into our products and platforms more accessible, more consumable so that we can unlock all of that value through, of course, our developers. You're going to see a lot of discussion about our developer cloud and some of the products and capabilities that we've put in there. You will see innovations from our platform providers, our ecosystem partners, software vendors, both large industry leaders as well as some of the new startups and the smaller companies that are you know really pushing on that leading edge of innovation.
That's amazing. It's such a big ecosystem, right? They're all coming here, Intel Innovation Live. Hey, I wanted to ask you, we just got done with our AI Global Impact Festival winners, and that was amazing, right? Technology education, especially in AI, is something that you're really passionate about. What are your thoughts on the next generation of talent coming in?
You know, this is probably one of my favorite parts about this event and the things that we do to unlock innovation in the world and use technology for good. When I, you know, get to hear from those next generation technologists and see the things that they're doing and thinking about and the way they're using all the ingredient technologies and innovations that we provide to have impact at scale, I get very excited. I'm looking forward to our little photo op with all the AI Global Impact Festival winners later today.
That's amazing, and I'm sure they're really waiting to see you also. Hey, Sandra, it was great, and back to you, Madeline.
All right. Good stuff, Ravi. Last year at Innovation 2021, Pat Gelsinger introduced the idea of superpowers and Intel's commitment to openness, open ecosystems, choice and trust. Moving on to this year, we expect today's keynote will highlight the progress Intel has made toward that goal, as well as the technologies that will be essential to developers with an increased focus on AI. We know there will be a few surprises in store as he and his guests share more on the evolving industry landscape and an expanded portfolio, of course, of secure cutting-edge technologies that will help developers build solutions that will change the world. Coming up, it's nearly time for Pat's keynote. Before we count down, we'll meet someone you're going to see a lot of in the opening keynote, AI ethics pioneer, Ria Cheruvu.
How many tacos did you order?
100.
The technology inside, the technology that moves the world. How wonderful is that? Intel.
Welcome back. We're coming to you live from Intel Innovation. The audience is filing in, and we're counting down the final minutes to the opening keynote. Our next guest graduated high school at only 11 years old. Now at 18, she has earned her Master's in Data Science from Harvard University and worked for four years here at Intel. Fun fact, when she's not writing code, she's writing poetry. Hear from this inspiring woman in her own words.
Hi, I'm Ria Cheruvu. I'm an AI Ethics Lead Architect at Intel, and this is my job in my own words. As part of my role, I'm responsible for tackling three primary categories. The first is ensuring that technologies are developed responsibly, both for the AI models and the mechanism sets around them. The second is helping influence our internal compliance processes to Intel. The third is helping influence the regulatory landscape around AI systems. I've been working in AI for about six years now. I started my initial foray into the field when I was 11 and getting my undergraduate degree in computer science, where I studied a lot of key concepts around big data systems, computational neuroscience, philosophy, and AI. I joined Intel formally as an intern when I was 14 and started to work on exciting applications around deep learning and reinforcement learning.
I'm now a full-time architect that has her Master's degree in data science. As a software architect, my goal is to be able to understand developers' pain points and be able to implement technologies that make some of these key concepts easy to digest and implement. My goal is to be able to help develop technologies and also empower the ecosystem, so developers are able to use these to make their life easier, but also are able to voice back and help contribute to these technologies. Communication is one of those essential items as part of the AI domain, and to me, my passion is around being able to digest some of these very complicated topics and then be able to present that to audiences. One of the coolest things about my job is being able to identify and dive deeper into very challenging problems around AI systems.
Collaboratively, we can start to solve some of these key problems, tapping into the expertise of some amazing perspectives around the world, and be able to identify how can we best prepare, design, and implement AI to improve quality of life. That's what I do for Intel.
You heard it in her words. Now we're here with Ria live. Ria, you are not only a huge inspiration, but also Intel's AI Ethics Lead Architect. What personally excites you about Intel Innovation this year?
Oh, thank you, Ravi. I think what personally excites me is being able to get exposed to a lot of this talent and learn from all of these different teams and showcase their awesome capabilities and performance.
That's amazing. Now, which demos are you looking forward to seeing the most during the next two days?
Definitely the ones corresponding to AI. All of them are super interesting. I'm really looking forward to being able to speak to those, learn more about them, and assimilate all that.
That's amazing. Now, I gotta tell you, I've been watching you and some of your talks. You talk a lot about curiosity, and I have a 12-year-old daughter, Angelie. What do you wanna say to little girls to inspire them to get into technology, to be curious? What are your words of wisdom for them?
Oh, absolutely. I think curiosity is, like, that central part of being able to pursue different learning opportunities. I'd say the first step to curiosity is having this. You're able to understand what you really want to explore and then go ahead and dive right deep into it. You're able to figure out, "Okay, this is where I'm gonna go next when it comes to pursuing that curiosity. This is what happens if I fail," and having that kind of rock to stand on, that's what I'm learning currently is the most important thing to pursuing your passion.
That is amazing advice. Amazing advice. This is your first Innovation event?
First Innovation event.
In person, how are you feeling about the energy and the buzz in the room?
Oh, feeling great. I'm just, like, learning all of the new things. I feel great. Definitely the team has provided so much support. Yeah, learning along the way.
Awesome. Well, great that you're here. Excellent to be here with you. Madeline, we'll take it back to you.
Thanks, Ravi. Man, she is impressive. Here's what to expect in our opening keynote. We'll have a host of demos showcasing improvements in both software and hardware. Expect to hear about some new tools that will accelerate development and also new ways to collaborate with Intel to build stronger open source ecosystems. Security is paramount, and we're looking to hear about ways to reduce risk and build software sustainability and catch a glimpse into the future with a number of surprise announcements. All right, Ravi, it is nearly time. Can you give us a sneak peek at what's going on backstage? I feel like it's a great time to be back there.
Madeline, I tell you, I love it back here, and I'm gonna give you a little bit of a tour. As you notice, there are some people in this room. Nobody's wearing capes, but I gotta tell you, they are our events and supply chain superheroes doing all of the things that have to be done to make all of this work and come to life. I'm gonna kind of bring you back, and I'm gonna bring the audience back. I'm gonna point over here. This is what we call demo land. These are the folks that are enabling all of the demos and the showcases, and the technology at the event this week. Lots of work. My man, Art, over there making it happen with his team. Yeah, lots of great stuff going on over there.
This is what I consider the center of the universe. This is our live stream crew over here. If it wasn't for these guys, you wouldn't be seeing this pretty face. These guys are making it happen. Okay. Then we've got Video Village. Video Village, they're the ones that are putting together all of the content to bring it to stage and bring it to the show. Now I'm gonna take you up on stage. Watch your step, guys. Oh, wait a minute. Look what I found here. Can you see this? This is Programming the 80386 book that was co-written by Pat Gelsinger. Okay, come on. Follow me out here. Let's see what's going on outside. Wow, look at that. Look at those smiling faces. Are we excited, everyone? Look at that.
Glad to be back here in person at Innovation 2022. Back to you, Madeline.
Thanks, Ravi. Oh my goodness. I love seeing all the waves and the smiling faces out there. It's obvious we have the heroes there. We have the developers here. The execs are here. The special guests are here, and of course, you, our online audience, are also here. Everyone, the time is here. Let's take it to Innovation 2022.
Please welcome Intel CEO, Pat Gelsinger.
Welcome, all. Good morning, good evening. [Foreign language] Shanah Tovah to all of our friends who are celebrating Rosh Hashanah, and our Israeli talent just play a huge role for Intel and Mobileye and our teams there, spending the holiday with their family and friends. To the rest of team Intel, we are here, and we're gonna have a great event together now. It's wonderful to be back. What do you think? In person. Yeah. In person, in our backyard here in Silicon Valley. A beautiful day today, so let's get started. You know, as the host of Innovation 2022, I have to affirm Intel's deep, unwavering commitment to an open ecosystem. Last year, we said we'd innovate on future technologies in the open. We would offer you choice. We'd help drive industry shaping standards, and as we said, we have to make them more secure.
These platforms have to be things that you can trust in and build your platforms and businesses upon. Over the next two days, we're gonna.
Show you our progress, but also look into the future in these areas as well. You know, we have a ton of news for you today in hardware, software, chip design, security, you know, new developer platforms, and everything the industry has to offer. We're gonna geek out. Did you get it? Right. Okay, all you geeks in the room, if you don't get it, right, you know, go back and check your ASCII coding again, right? This is geek in ASCII, right? So, we're gonna have some expected guests, but some unexpected guests as well. We're gonna have a great show for you this week. You can think about this as a keynote in two parts, right? You got me today, and you got Greg Lavender, our CTO, tomorrow talking about many of our software first and developer-focused activities.
For those here on site, you know, join me, join us in our Innovation Zone, where we're gonna have some geeky demonstrations, get to look firsthand. For those of you online, wish you were here, but the online experience is gonna be great. Stay on afterward for expanded coverage, and you don't wanna miss it. Let's get started. You know, we're in this new era, and it may not be quite post-pandemic, but we're starting to see our way to the other side of it. Through this period of time, we have seen the increasingly critical role that technology plays in every aspect of human existence. Everything is going digital. How we work, how we learn, how we rest, how we care for each other, how we worship, how we live in society and in community together.
In this digital era, we continue to witness the magic of technology, our ability to push forward with innovation, discovery, and growth. This is driven by what I like to call the superpowers, these foundational technologies that are profoundly shaping how we experience the world, and it creates this bridge from the analog, from the physical to the digital world. It enables us to simulate, to create the digital twin of everything that we do. You know, one of those, compute. Since the 4004, we've just been pushing compute into everything. Connectivity, everyone and everything is becoming connected. Infrastructure with unlimited scale of cloud, but increasingly with edge, unlimited reach for capacity as well as low latency and bandwidth. AI intelligence everywhere, turning this infinite data into actionable insights.
You know, I've been framing these four superpowers for a while, and then in conversation with some industry peers, and in particular, a conversation with Thomas Friedman, the Pulitzer Prize-winning, you know, New York Times author, he convinced me I was missing one. Sensing. You know, everything that we compute and connect is now becoming increasingly able to sense. We see objects and even be able to see things that you can't see, right? Hear things, know exactly where you are, and increasingly be able to feel, taste, right? All of our senses. I want an augmented reality where all of my senses become digitally enhanced as well. Even my infirmities, like here's my hearing aid. Yeah, you know, part of our family, right? We lose our hearing, but now. What did she say again?
I know more about you than you think I know. Even my disabilities become sensory devices to make my every day better. Each one of these five powerfully, individually, you know, just changing. But as you think about it, more sensing means more compute. More compute means more connectivity. More connectivity, more infrastructure. More data over that infrastructure, more AI, which leads to more use cases. Right? They're reinforcing, accelerating, empowering each other to unlock even more powerful possibilities in the future. You know, and, you know, it also changes how we think about our business as well, you know, and how we think about chip design and, you know, the role that we play in this industry. With that, we launched Intel Foundry Services, a key part of our IDM 2.0 strategy to help meet this growing demand for advanced global semiconductors.
We've committed to a strategy of helping to build a more balanced and resilient supply chain for the world. At its heart, of course, is Moore's Law. You know, and for decades now, I've been in the debate, you know, is Moore's Law dead? The answer is no. With advances in transistors, with RibbonFET, advances in power delivery with PowerVia, right? With breakthroughs in lithography, with High NA lithography, the core of semiconductor manufacturing, with advanced packaging technologies, we aspire from today about 100 billion transistors on a single package. By the end of the decade, 1 trillion transistors. 1 trillion transistors in a single package. We are on schedule or ahead of schedule on this audacious strategy that we have said five nodes in four years. Normally takes two years for a node. We're gonna do five in four years.
Are these Intel folks crazy? No, we are just torridly moving to the future. Our 18A, right? You know, the key one where we'll assert as unquestioned leadership, you know, the PDK 0.3 is in the hands of developers today, and we expect our first test chip taped out against that before the end of the year. We will not rest until the periodic table is exhausted. We will continue to be the stewards of Moore's Law into the future. You got it? Alive and well. Of course, to be a great foundry, you gotta be a great wafer foundry, and we're gonna continue to work to grow that semiconductor demand, build more capacity, work with our customers as a foundry. We've announced our intention to acquire Tower Semiconductor.
Combined with IFS, we will have a global diverse end-to-end and one of the richest portfolios of differentiated technologies available today as a foundry provider. Our vision is much greater than that. IFS will usher in what we call the systems foundry era, and as the focus moves from system on chip to system in a package, and we see a systems foundry having four components to it. First, you gotta be a great wafer foundry. We're not debating that. Gotta be good at that. Second, right, you know, it has to be great at packaging. Even Gordon Moore, when he wrote his original paper on Moore's Law, he foresaw this day of reckoning where we'll need to build larger systems out of smaller functions, combining heterogeneous and customized solutions.
With 2D, 2.5D, and 3D stacking, you know, this gives us the ability to increase the number of transistors per device. Third, software. As you're gonna hear from Greg tomorrow, he calls it software, the soul of the machine. You know, we'll take these stable software interfaces that allow you to accelerate your product development and delivery. You know, open source software stack, the key software components that unleash the different chiplets, accelerators and abstractions. We're, you know, increasingly gonna make that common through our oneAPI solution. As we like to say at Intel now, software defined, silicon enhanced. Software comes first. The fourth, a vibrant chiplet ecosystem. Just as Intel has played this pivotal role in standards, you know, like PCIe and Ethernet and Wi-Fi over time, you know.
My granddaughter, when she plugs in a USB stick into the computer, she says, "Thank you, Papa." Yeah. You may like me, but I care more what she has to say. Once again, we're gonna lead an ecosystem to create the standardization of a chiplet, you know, architecture. We'll call it the Universal Chiplet Interconnect Express, or UCIe. We already have more than 80 companies who are participating in this creation of this open chiplet environment. You can imagine something like this, "Hey, I'm getting a few of my chiplets from Intel, but, you know, hey, I have a few of these pieces that I've also developed on TSMC. You know, maybe my power delivery components are coming from TI, and I have some o f my I/O chips that I'm doing on GlobalFoundries. Bring those together.
Of course, you know, "Intel with the most advanced packaging technology, assemble those together for me and be my supply chain manager." You know, we're gonna bring all of those pieces together, enabling what, you know, we'd say is this next generation of the chiplet ecosystem. Let's just hear from a few other voices that are joining with us in this vision.
At TSMC, our goal is to provide an open silicon platform to unleash customers' innovation. That's why TSMC fully supports moving the UCIe ecosystem forward.
The future is chiplets. This is why Samsung supports the UCIe goal to establish an ecosystem of chiplets that will drive the next era of heterogeneous computing. Thank you.
With Intel, Samsung and TSMC, you've just heard the three largest silicon manufacturing companies committing to a common direction. As part of our IDM 2.0 strategy, you know, we're committed to investing in the fundamental capabilities to enable this ecosystem, but also the next generation of talent as well. Part of our $1 billion IFS Innovation Fund is directly investing in companies and startups and disruptive technologies. We've had phenomenal response, even in the early days of forming this ecosystem. We've already made our first investments and many more in the pipeline to come. We also have this, you know, desire to expand the talent much more deeply, and we have our university shuttle program that we've been running with a select set of public and educational institutions to enable them to have modern technology for their classes, training and talent development.
I've given our team an audacious goal to scale our shuttle program, the passengers, the number of chips that we're doing, 100x. Did you hear that again? Right? A 100x increase in our shuttle capabilities. You know, professors, researchers, students, you know, all of them being trained on the latest EDA tools and having access to shuttles that they can be sending their chips to manufacture on the most leading process technologies in the industry. I think of this just like when I did my first DRAM design and sent it to MOSIS for manufacturing. You know, 40 years ago, but we want to build that talent pipeline for the future, and we're gonna work with these institutions to radically increase the semiconductor talent flow of tomorrow. This is our commitment to the future of the semiconductor industry.
Switching gears, you know, another passion of mine has been in this area of graphics and accelerated computing. We've been delivering integrated graphics for decades. We've also known for decades that high density computing, you know, high performance vector, matrix, the highest bandwidth requires a different solution as well. You know, this area of HPC has ushered into AI, machine learning, and core graphics technologies like ray tracing. You know, and personally, I've been on this journey for 16 years. When I left Intel 12 years ago, you know, it was the only major not done on my list of things that I wanted to finish. Well, I'm back, and we're now gonna get it done. With that, I'm delighted today to announce our GPUs. We got not one, but three of them to describe today.
The first GPU is our Intel Data Center GPU Flex Series. This product, pixels are growing exponentially for a range of workloads, visual cloud tasks, media transcode, cloud gaming, inference and training in AI. This is the industry's first hardware-based AV1 encoder for a data center GPU, delivering more than 30% bandwidth savings or more streams per platform, however it's monetized by the cloud provider. Open and full-stack API access. Today, Flex Series is supporting oneAPI, of course, but enabling frameworks like TensorFlow, OpenVINO, and PyTorch as well. GPU Flex. Of course, the first cousin of big AI in data centers, high performance computing. This is our Ponte Vecchio, our flagship high performance GPU that uses our most advanced IP, really setting the pace for our most advanced packaging technologies, petaFLOPS of performance.
Together, Ponte Vecchio, along with our Sapphire Rapids HBM, our 4th G en Xeon, right? Is the computing brains behind the Aurora Supercomputer, and together, we're delivering these blades right now to Argonne National Laboratory to build the world's highest performance supercomputer. Where did it all begin? You can't talk about GPUs without talking about gaming. You know, and for a long time, you know, we've been seeing that the average price of GPUs is right in this, you know, $200-$300 range. What's happened the last few years? They've just gotten super expensive, and we don't think they need to. Today, we're about to fix that. You know, and we are, with gamers, delivering and hearing the complaints of high prices, so you should be frustrated because you are losing out as the gaming community. Today, we're fixing that.
Here is the Arc A770, right? This is our top end gaming desktop GPU packed with features, delivers 65% better peak performance in ray tracing versus competition. We've come a long way. Let's take a look at the Arc A770. Yeah, I'm excited too. I'm powering mine up, hopefully this next weekend. It'll be available October twelfth, starting at $329 price point, and the cards are on the way to our reviewers right now. I hope you're excited. I know I sure am. You know, you're all here at Innovation because you share my passion. You know, the shared passion to continue to innovate, to continue to use technology as a force for good, to serve the technologists on the front lines, and to use technology to solve today's biggest problems.
We've been around for a little bit over 50 years, and we're the company that put silicon into Silicon Valley. I remember as a young technician, when I walked into Intel, I was so impressed by the technology leaders that they were there. I thought I had entered this sacred place going into the temple of technology, where everyone was maniacally focused on what technology could be and how it truly could change the world. With that, I want to bring on stage a young, bright technologist who embodies that passion to harness technology as a force for good. Ria Cheruvu, right, graduated high school at 11, started working at Intel as an intern at 14, the same year she received her undergraduate degree from Harvard.
She's a whopping 18 years old today, the same age I was when I started at Intel, and you can imagine on stage I am now the underachiever. Please join me in welcoming Ria to stage.
Hi, Pat. Thank you. It's wonderful to be here today.
Well, thank you, Ria. You know, like I say, boy, you know, it's not often I feel like I am just an underachiever starting at 18. Man, you've done incredible things already.
Thank you.
Tell us what you're working on now?
Oh, absolutely. Right now, I'm an AI ethics lead architect. My goal is to develop technologies that help make AI more responsible, trustworthy, and ethical.
Well, you know, you and I have talked about the importance of taking on challenges, you know, and not being afraid to fail. What do you think?
Absolutely. I mean, I'm ready to take these on, head on and solve them, so.
You know, let's talk about this vast community, this vast universe of tools and technologies we're making available to innovators and developers like you, Ria, and how we wanna continue to provide easy, affordable access to the latest stuff. We wanna help you get your hands into it without downloads or tough hardware setup. To that end, right, and you'll hear more about this in Greg's session tomorrow, right? The Intel Developer Cloud, giving developers access to new and forthcoming hardware solutions for pre-launch development and testing.
Absolutely. I think my peers are ready to get their hands on this technology and start innovating with it.
Yeah. We're expanding our Intel DevCloud experience. We're not trying to compete with the Amazon, Google, and Azure of the world, but to win your hearts and minds with early access of the latest great technologies that we're working on. Today, the Intel Developer Cloud is limited beta to, you know, select and pre-qualified customers, but our objective is to expand that significantly to developers over time for their platforms and to enable the scale of this service in the future.
Absolutely.
As part of this trial, you know, we're putting the latest and greatest stuff out there, you know, and you'll have access to our latest technologies, you know, fourth gen Intel Xeon, also known as Sapphire Rapids, to Habana Labs Gaudi2, right? To our advances in IPUs and, you know, our NEX technologies. This is just the beginning, and you'll get updated this week on the full set of things, because we want developers to have access to the latest and greatest.
Oh, awesome. I know that there's a lot of AI developer kits, software kits, reference implementations out there for developers to get their hands on and use right away. Yeah.
Today, what I wanna do here, Ria, if you're okay with this.
Okay.
We're gonna set up a handful of challenges, and you're gonna help me solve those challenges today. Along this journey, we're gonna meet a few other technologists, and we're gonna tap into the open ecosystem, choose the right tools, and we're gonna use it to build responsible solutions to today's problems. You game?
Absolutely. Let's get started.
Our first challenge, speed up development cycles. One of the great debates over the history of technology has been this open versus closed, proprietary, protected gardens versus the open marketplace. Since the days of Grove and Gates and Jobs, this, you know, has been pervading the industry. We believe that this idea of the persistent, expensive, closed environments. We are the champion of that open marketplace, and we are back. John from our demo team is here to help us talk a little bit more about this challenge.
All right. Awesome. Great to meet you, John.
Thanks, Pat and Ria. The challenge is really speeding up development cycles, and I have a tool that I think can help us with that. The Neural Coder, which is a new feature of the Intel Neural Compressor, is a one-click, no-code solution that removes hard-coded CUDA references from our deep learning scripts. It optimizes those scripts for performance, and it benchmarks those optimizations for the best possible deployment options. Let's take a look at how easy this really is. On the screen, you can see we have an image classification workload running. We're going about 300 frames per second with that. Ria, you wanna try the Neural Coder and see what kind of impact it can have?
Absolutely. All I'm gonna do here is go ahead and click the Run button and confirm, and is it that simple?
That was pretty simple, right? Let's take a look at the Neural Coder output. The Neural Coder has now optimized our deep learning model from FP32 to INT8, and it's really leveraging the Xeon Scalable processor with Intel Advanced Matrix Extensions.
That's fantastic. What's going on with the performance?
Yeah. Let's take a look at the image classification workload again. We're now running at about 3,000 frames per second, which is a 10.8x speed up, really leveraging that built-in acceleration with the Xeon Scalable processors.
That's looking good, John.
Yeah.
All right. Perfect. I'm wondering here then, with this type of one-click optimization, everything's good to go and all easy to use.
Yeah.
Perfect.
Maybe we can start talking about security now?
All right.
Well, you know, one of the things is that, you know, security, right, you know, enables the, you know, ability for us, right, to protect our models, to protect our data. With that, you know, how can we make sure all this model work that we do is actually useful and the data protected around it? Can you solve that problem for us, John?
Yeah, let's take a look at this over here. Let's make security a little bit more simple. Confidential computing with Intel SGX protects your data in use. Here we've downloaded a curated container, and we're gonna leverage the Gramine Library OS project, which, with a simple script and answering a few prompts, you can protect your data in use. You wanna give this one a shot?
Yeah, absolutely. All right, this is another one of those one clicks. Gonna go ahead and click here and cycle through these defaults.
Answer a few prompts here.
Right. Says it's being created.
Awesome. All this data that we're using over here, you can now see on the screen that we've protected that data.
Amazing. This is confidential computing. My container is SGX enabled. I've got protections for my data, and I'm all set when it comes to that.
Great. Optimizations and security.
All right.
You know, very good. You know, it's great to see that we're starting to get real traction from the open source community to take advantage of these confidential computing capabilities. You know, we're working to enable a holistic view of confidential computing. You know, Project Amber, which we'll talk about tomorrow, a third-party trusted attestation service, which Greg will talk about. You know, we saw the full pipeline today, right? Coming into place, right? What we're doing with AMX as part of our Intel 4th Gen Xeon to accelerate, right? On the CPU as well as on GPUs, the capability, right? You know, making it complete and protected with SGX, but we're also working on tomorrow with TDX, our trusted domain extensions, which will additionally provide secure enclave at the VM level without requiring application change.
This helps everyone to have more trust in their AI and in their data. Ria, you know, I think we've covered and we can call this one complete in our challenge. Okay. Let's get to our second challenge of the day. You know, we saw that we can do them faster, but how do we make them more relevant to the marketplace, these AI models that we're doing, you know, without requiring an advanced degree in AI modeling? Do you think we can do that?
Absolutely. Let's go for it.
For this demo, we're gonna come over here to this side, and joining us on stage is Yannis from our computer vision and AI team. Yannis.
Nice to meet you.
Nice to meet you.
Thank you.
Today we're going to talk about AI model development, and this is still a very slow and expensive process. Even then, only a fraction of the models make it to production. Why is that? There are several reasons, but the main reason we see is that the main experts are left out of the loop. When we hand the model over to them, it may have drifted, making it unusable.
Oh, absolutely. I've seen these issues all the time, accuracy drifts, poor inferencing quality, and then there's also problems with the outputs of AI systems. I'd also like to see a lot more collaboration between domain experts and data scientists. Really drive those collaborations, get to higher quality models faster. You have a solution here for me?
I have good news for you.
Okay.
Intel has built an intuitive computer vision platform that enables anyone on enterprise teams to collaborate and build models very efficiently. Let me show you an example. This is an image of a coffee plant which is behind the $400 billion industry.
Well, you know, coffee's pretty important to me, right? You know, enough coffee, all things are possible, Yannis.
Yes. We say developers turn caffeine into code, so. We want to help those coffee farmers to use AI to build models that can predict the maturity of the crop. This way, they can maximize the yield. Let me show you how you can build a computer vision model to do that. We have this project here, and I have already uploaded some images. Let's go and open it.
All right. A question here. I can only see a couple of images, but in a typical computer vision workflow, we'd probably need tens of thousands of images for this step.
Well, that's a thing of the past. With this software, we are using a technology called active learning to drastically reduce the amount of training data that a system needs to train. Let me go ahead and open this image here. I will select one of our annotation tools, and I will throw a bounding box with my stylus around this yellow fruit, which is immature, this orange one, which is semi-mature, and this red one, which is mature. When I submit those annotations together with some other annotations we did earlier on, the system will start a new training round.
All right, perfect. What happens after training?
Well, it's gonna take a few minutes for the training to complete. For now, I have already created a project. I will go and open a previously unseen image. Let's do that. What do you think, Ria?
Okay, this model's doing really well on this image. I am seeing a couple of erroneous and missed detections, though.
You're right. You can improve this model by correcting those missed detections. Would you like to try out?
All right. On the fly. Okay, perfect. I'm gonna start with deleting this little leaf annotation, and then I'm gonna go ahead and navigate to the quick selection tool. This one looks overmature, and this one looks mature to me. All right.
A coffee farmer already, Ria. Or an AI expert. Which is it? I'm not sure.
Probably the AI expert because yeah, I mean, I definitely need that type of expertise, though, to be able to do this type of labeling. This is how we can get them in the loop and then make sure that we converge to better models quickly and faster, so. Right?
The most important part is that once your model is trained, it can be optimized with OpenVINO and deployed across different types of hardware at the edge or in your data centers.
Amazing. Pat, I love this technology.
Well, Yannis, thank you for joining us and showing off. This is incredible progress.
Thank you.
Thanks a lot to the team. Thank you.
You know, it's a real demonstration of how Intel is building practical, useful ways to deploy AI. Today, we are launching Intel Geti. Powerful AI for everyone. What you've seen here is this new computer vision platform lets anyone in the enterprise teams build computer vision models rapidly. I mean, quite amazing. Big data? Oh, no. We can do it with small training workloads, and this is now commercially available in Q4 of this year, enabling enterprises to digitally transform and mainstream AI easily into their operations. A true breakthrough. Aren't you excited? Okay. There we go. Now we have, right, how we create the AI models, but of course then our deployment of AI, and this is where OpenVINO has been extremely well accepted by the community.
Running on this range of Intel hardware, a fast deployment creation, deployment to complement Intel getting hundreds of thousands of developers using OpenVINO today. We have an in-market example to show you here today. Please join me in welcoming VP of Operations of PreciTaste, the vision AI food management platform, Hauke Feddersen. Hauke?
Hey, Ria. Hey, Pat. I hope you're hungry because the next live demonstration will take us into a restaurant.
Ooh, okay.
We help professional kitchen crews by eliminating manual estimation tasks. Our valued customer, Chipotle, focuses on enhancing the customer and the crew experience. Vision AI tells them how much inventory they have at hand and what to cook, when to always serve the freshest food and to never stock out. All of this is running on the edge in the restaurant on small form factor edge devices like that one. We chose OpenVINO because other companies' inference-only edge solutions always just focused on the inference and neglected the CPU intensive workloads that we need to run in the background, like databases.
Okay, that sounds great. One of the first problems, though, that I'm encountering and I'm thinking about that you may have experienced as well is, you know, how do you know how much compute you need and whether or not to use the Intel CPU or the GPU?
Great question. OpenVINO opens up a huge variety of devices for us to choose from. When we pick a device, we benchmark it to verify that it has enough performance for this and future use cases, and we deploy it. It's actually the auto optimization plugin in OpenVINO that makes the decision whether to run our model on the CPU or on the integrated GPU. The best part is, with each generational upgrade, our customers get so much more performance out of one edge device. The generational jump from Gen 11 to Gen 12 took us one afternoon, and it resulted in 30% more frames per second for our application.
Amazing.
Hold that thought. I mean, that's pretty powerful. Essentially, OpenVINO allows you to not worry about the underlying hardware, right? allows you to be, you know, taking, you know, advantage of these capabilities. It's like having a world-class performance engineer standing by your side as you deploy.
Absolutely. Let's see it in action. In the restaurant, we have those Intel RealSense cameras mounted above the make lines, seeing the food in real-time all the time. We use the OpenVINO to detect the corners of each of those pans for a dynamic region of interest. We use OpenVINO to classify the ingredients to know what we're looking at, and we use the stereo sensing from the RealSense to tell us how much we're looking at.
Right. That's a lot of models that are in, like, part of a workflow. How is OpenVINO helping you here with, like, this much compute?
We're running multiple applications on just very few of these edge devices in the restaurant, from what to cook when, to digital prep, to freshness monitoring, and order accuracy verification. For that, we need tracking performance. Careful quantization in OpenVINO gave us a speed up of 2x. We're now running live inference from two vision streams on just one edge device.
All right. Your performance engineer has a question, or your temporary performance engineer. If we switch around those pans, how are the depth sensors gonna react? Give it a challenge.
Let's do it. If I take out one pan.
All right.
... and shuffle them around a bit, we can see that the AI is orienting itself again and knows exactly how much food is available. If I start serving this delicious salsa. Dig in, dig in.
I like salsa.
The inventory is depleting, and the crew is being asked to replenish so that the restaurant never stocks out. Chipotle is using this computing one-stop shop to scale their amazing operations. Don't take my word for it, hear it from Chipotle.
You'll see it now.
Hey, Pat. Hey, Ria, and everybody at Intel Innovation. By utilizing Intel technology in our innovation center and test markets, we've seen a real improvement in operational efficiency. The combination of the full Intel architecture and OpenVINO software has alleviated manual tasks for our crew and given managers the tools they need to make informed in-the-moment decisions. By deploying Intel's RealSense and OpenVINO technology, we created a powerful and versatile AI solution that addresses our unique needs while connecting the entire infrastructure of the kitchen. We're excited with the results we've seen in our test market and are aiming to fully deploy Intel technologies across Chipotle restaurants in a large market in the coming months. I'd like to thank you all, the true creators and innovators in this space, and hope you have a great show.
Well, thank you to Curt. My family loves Chipotle, and now it'll be live and have Intel inside Chipotle, so we can go there more often.
Absolutely. With this delicious demo, I can say this challenge is complete.
All right.
All right.
Let's move to our next challenge.
Yep.
If you would. Game development. You know, modern games are complex. You know, they have, you know, interactivity, lifelike graphics, complex characters, story development, immersive experience. You know, and these are big, expensive development and production operations to bring them to market. You know, this is a lot of work to get done. To talk about that, I've invited on stage today a world-class game developer, Inflexion Games, right? Who's striving to bring a realistic, immersive gaming environments onto the PC. Please help me in welcoming Sarah Mainwaring to the stage. Sarah?
Hi. Good to meet you.
Hi, Sarah.
Hi, Ria. Good to see you.
Thank you.
All right, I'm from Inflexion Games, and we are creating Nightingale, which is an online multiplayer survival crafting game. When we're working with a multiplayer game that's online, we get networking bugs, and these can really tie up our development schedules when we try and debug them. While you know, we're very good at doing this, sometimes we don't have any shortcuts. When we are doing our debugging for these issues, we are connecting multiple clients on a single machine in the editor, usually with a test level. Here's the test level I wanted to show you today.
Oh, that's beautiful. Looks immersive. Am I hearing birds?
There is a little bit of birds coming out, yeah.
Yeah.
This isn't the usual test level that I would use to debug anything in. This is a real Nightingale player level, running an Unreal Engine 5, on this rig. Usually, we would be working in a test level that's very small, contained, very empty, so that we can really drill down to what the problem is. Sometimes that's not possible. Sometimes there is a bug or an issue which only shows up in player-facing levels, and then we run into problems because we have to run these very high fidelity, dense levels directly on our work machines on a game which is only optimized to run one client on one machine at any one time without the editor even attached.
Absolutely.
You can see that we've got some pretty good performance going on here. It's really nice and smooth. It's looking real beautiful. On my workstation, you know, back at work, this would be what we would be getting in the editor. We won't be able to connect more than one, maybe two clients before the frame rate just drops off. On this wonderful workspace, we've not only got one going. Oops, excuse me. Give it a little punch there.
No problem.
Oh, well. Sorry, my keyboard is not working.
No problem.
Yeah, no worries. Let's see what I can do. Yep, my keyboard is not working, unfortunately. Oh. Okay, switching over to some backup images, but you'll see. We switched over to some images here.
It's real.
It's real, yeah.
Are they?
Right.
Development problems. What you can see here.
No, I think, Sarah, we found the bug.
Yeah, we found the bug.
Okay, good.
What you can see here is that.
There we go.
... we are running four.
Mm-hmm
of these instances all at the same time with the same frame rate. It performs wonderfully. Actually in back, when I had a bit of downtime, I tried eight, and I was able to run eight of our game instances, which are only optimized for a single on the client at any one time. Yeah, just really impressive tech.
Absolutely. I think peripherals aside, this is exactly what we're starting to get into. Game development's just gonna get better and better. Sarah, to your point, to highlight a cool feature here, I think hybrid architecture is definitely what developers can use here to make their lives much more easier. You can hide the windows that you don't really need so that your OS and, you know, your different platforms and applications are able to take advantage of what they do need in the front view and hide or put away the stuff that you don't.
Yeah.
It's a beautiful test environment, and it's a beautiful use case.
Yeah. Being able to connect four, eight, however many we try to do clients at the same time, just speeds up our ability to debug and get to these nitty-gritty problems that are only showing up for our players. It's a big deal for us to use Intel technology for this.
Absolutely. Definitely as part of that hybrid architecture performance and efficient cores or P-cores and E-cores, that's definitely gonna come in handy for these types of tasks.
Yeah. This is in editor, and we've prepared a little bit of footage to see what Intel technology is giving Nightingale outside of the editor. Can you run the footage?
All right.
Let's see it.
Gonna roll it. Epic. I caught Pat having a jump scare, but I am so excited to be able to play that.
Yeah. I look forward to you all getting to try it soon, hopefully.
All right. Thank you, Sarah. This was awesome.
Thank you.
Thank you, Sarah, for joining us. Exciting accomplishment with the game coming out, and can't wait, 'cause I know end users love it. Our platforms want to consume it and deliver it, and we're committed to making all this real. Thank you so much, Sarah.
No, thank you, Pat. Thank you, Ria.
Thank you.
As you just saw, the future of gaming and creative immersive experiences like that relies on performance CPUs and GPUs. With that, today I am thrilled to reveal the 13th Gen Intel Core desktop processor family. Formerly known as Raptor Lake, the flagship i9-13900K is coming next month. This is the fastest, best desktop processor ever built in history, and it takes advantage of our hybrid architecture, both P-cores and E-cores. You know, we've enhanced the P-core, so it has the best single-threaded performance ever. More L2 and L3 cache. We've also increased the number and improved the E-core capabilities for compute-intensive workloads and workflows. Together, the E-cores and P-cores work together in a hybrid architecture to deliver the best multithreaded performance.
It has blazing clock speeds of up to 5.8 GHz, and the i9-13900K is the best gaming, streaming, and recording experience ever. Make no mistake, it's the best CPU, but it's also the best platform. Our Core processors deliver on the platform value propositions, state-of-the-art wireless connectivity, advances in memory technology. As always, we are building and leading open platforms with and for the industry. Whether you're a beginner or an expert, the Intel 13th Gen Core delivers the world's best overclocking experience as well. You know, early next year, we'll release a SKU that, for the first time ever, hits 6 GHz out of the box in limited volumes. An industry first, a huge milestone for client computing. I remember like, just like it was yesterday, when we first crossed 1 GHz. Man, you all are blowing me away.
You know, 6 GHz, what a milestone. Congratulations to our team. Of course, we're not stopping with the K SKU. We're gonna give you the full range of PC segments, including mobile. We expect over 50 processors as part of the Raptor Lake i9 family, or the 13th Gen family. We have over 500 13th Gen designs.
I'm back, and I think we should definitely get Hauke that version.
Yes, I think so. Our game is definitely gonna take advantage of this. Given the sheer power and unique hybrid architecture that comes with the 13th Gen processors, I think we can say with confidence the gaming challenge is complete, Ria.
All right. Thank you.
Let's move into our fourth challenge, and this one's for content creators that need a creative boost.
All right, let's do that.
I think we need to commemorate our time together today, Ria.
All right.
With that, why don't we do a movie poster?
Okay.
Right? You know Pat and Ria's greatest hits kind of poster. You know, have you ever seen any of the text-to-image creative platforms?
Well, I've definitely seen it. I could even say you're asking the master. I have tried so many different Stable Diffusion models. It's constantly growing, but definitely something that only advanced deep learning enthusiasts can get their hands on. Very hard for content creators to be able to use that technology right away.
Okay. Well, AI is unleashing this wide array of new opportunities for the creative process. John is back on stage to show us some amazing technologies from Habana Labs using Gaudi2. Maybe we can help create our movie poster, John? Let's do it.
All right.
You mentioned text-to-image conversion, and we're running a latent diffusion model on our Gaudi2 deep learning accelerator. Pat, Ria, you wanna start the creative process?
Well, how about movie poster?
All right. I'm doing it. Stakes are high.
The Gaudi2 is generating 16 images here in just under 3.5 seconds.
Fantastic.
Okay. Well, can we go a little bit further? What do you think, Ria?
Absolutely. Let's do superhero.
Okay
Poster on the fly. Gonna type this out here. Let's see.
Okay, that looks better. Okay.
All right.
you know, what do you think? Should we take it a little further?
Yeah, absolutely. Let's do something a little bit more cool.
Maybe, you know, male and female superhero movie?
Yep, absolutely. In the style of Art Deco. That's gonna give us a couple of cool images that the model is trying to generate pretty much real time.
All right. I think we're looking pretty good here, John. What do you think, Ria?
I think it's great. The speed is phenomenal. It's giving me a lot of inspiration for different ideas that I can use. You know, definitely some improvements, but and yeah, I think it looks great to me, John. Thank you.
Yeah. I can't wait to see designers use this and speed their creativity.
Absolutely.
You know, John, I think what we've seen here is the incredible power of the next generation Gaudi processor. Here I have the Gaudi2 accelerator. This product comfortably outperforms the A100, ResNet-50, BERT models, the June publication this year of the MLPerf benchmarks. You know, this model we're running is latent diffusion on PyTorch. What do you think, Ria?
Absolutely. Those words, the machine learning models, they bring joy. I think this is fun. We're definitely done with this challenge, and looking forward to seeing what comes next.
Okay. This challenge complete.
Okay. I think that was a great challenge, but I'm also encountering a little bit of a difficulty, Pat. Something that I really struggle with is you need a lot of devices in order to be able to do this type of content creation and being able to, you know, train your deep learning models, and those devices don't even talk to each other.
You know, I think, Ria, that sounds like a good challenge to me that we can attack this morning.
All right.
You know with Centrino, integrated Wi-Fi, we created an industry. two-in-ones and convertibles. You know, we've just been driving these unique experiences throughout our history. We start with people, what do they want, and design solutions for them.
Okay, perfect. Here's what I want. Maybe there's a solution out there for me. I need a solution to having to, like, lug around my very large monitor all the time.
Okay. Right, making it a lot easier. You know, I was in Korea a little while ago, and when I was there with our partner Samsung, they showed me a technology that I think might just solve that problem. I'm delighted to welcome the CEO of Samsung Display, JS Choi to stage.
Hey. Good to see you.
Thank you.
I saw this prototype when I was on a trip to South Korea, and I instantly saw the potential for our platform. I just wanted to let everybody geek out a bit with JS, Ria, and I. JS, what do you got here?
Yeah, it's great to be here. This is a small PC, right? Sometimes we need there's a bigger screen. Let's see. I show you the magic. Whoa. Oh. Oh.
Okay, my turn. My turn.
Okay.
Okay, I gotta play.
Okay.
Oh.
Okay.
All right. I wanna try. Phenomenal.
I'm sure this shows one great example of future trend of PCs. It's because I'm here today to demonstrate this demo. It's very important for Samsung Display to collaborate between ecosystem developers in hardware and software together to achieve this kind innovative form factor PCs for best user scenarios. Lend me of this one. Today, we announce world first 17-in slidable display for PC globally for the first time.
Thank you.
Wow.
Yeah. This device will satisfy Ria's need for larger screen and also its portability as well. With flexible display, we implemented sliding technology rather than foldable. Foldable is gone. It's just beginning, right? With this slidable flex and software collaborations, we expect to see better and best user scenarios of PC near future.
JS-
Thank you.
Thank you for joining us here. Right, as the leading provider of OLED technologies, there's many decades of partnership. You know, this just convinces me we got a lot more coming between us. Thank you for joining us, JS.
Okay. Thank you.
Thank you.
Thank you, Ria.
I'm assuming I can keep it.
Again, continuing new form factors. We create new SDKs to enable those and find new experiences for our customers and community. Speaking of which, earlier this year at the CES show, we announced our acquisition of Screenovate, and this is now coming forward as Intel Unison. To show an example of Intel Unison, please join our demo geek, Craig, coming to stage. Craig?
Look at that. Hello, Ria.
Wow.
How you guys doing today?
Doing great.
I'm here to show off some amazing things that we're gonna be doing with Unison.
Okay.
For the first item, might as well pair some of this amazing software that we have in Unison, ooh, with some of this amazing hardware from Samsung Display. Okay, beautiful. Really what we're seeing with Unison and the promise of Unison is to bring all of our devices together, regardless of ecosystem or how they move and how they interact. All of our smart devices, again, really working for us in the way that we want to. Just like you said, Pat, with Intel, we're trying to solve the world's largest problems. Well, Pat and Ria, I have a lot of problems. Not those type of problems. Device problems.
Something like this, I have my great little Evo PC here, and with just a quick, nice little tap, I'm gonna go ahead and send it on over to my display. Oh, let's see here. Here we go. Ooh.
Ooh.
I've extended my display to this nice little 13-in to 17-in device.
Nice.
Which helps me with a lot of my problems. I like to have a secondary screen for a lot of road warriors like us. I can take a little 13-in tablet that I have in my backpack and then turn it into a 17-in monitor. No back problems, no TSA.
You're telling me that could I connect, like, my iPhone to the device or?
Ria, that's exactly what I'm telling you.
All right.
All of your devices. That includes iOS and Android devices like this, all working together for the first time with our Windows PC. Ria, your iPhone will be singing and dancing and playing with your Windows PC. Never lose focus, get all of your notifications, all of your emails, all of your texts, and be able to bring that vision. That's the dream, right, guys? All of our devices working for us, and we're here to bring that dream to you with Unison.
Amazing. Thank you.
Thank you, Craig.
All right. Thank you, guys.
You know, Intel Unison is a new software solution promoting this open ecosystem approach, solving Ria's problem, right? You know, but also enabling the iPhone for the first time interacting seamlessly with the PC, all our iOS, Android devices. Coming with new laptops this holiday season from our partners at Acer, HP, and Lenovo. This is just the beginning of what we're gonna be enabling with Intel Unison. What do you think, Ria?
Absolutely. It's astonishing technology. You didn't say no when I asked if I could keep it, so I'm making some assumptions. I would say that this challenge is complete.
All right.
Thank you.
You know, our final challenge of the morning. Sometimes the biggest challenges we face in technology are at the pace of change. We can invent these innovations, but is the market ready for them? Sometimes the technology isn't ready for the market. For instance, my passion for silicon photonics and, you know, this.
Well, our vision is to fully enable our chips with optical capabilities such that our processors today, piece of silicon, a piece of optics, fiber going directly into that. The results of that will be much higher bandwidth, much higher communications capabilities at dramatically lower costs. Today, optics is a niche technology. Tomorrow, it's the mainstream of every chip that we build.
That was a video of me two decades ago predicting this moment.
Right. That was the year I was born.
Oh, Ria. You are killing me. Sometimes innovation takes a little bit longer than we want. You know, sometimes a decade to make a week of progress, sometimes a week to make a decade of progress.
Absolutely. I love that quote. I understand it takes a lot of hard work and talent to make these ideas into reality.
Yeah. At Intel, we invest for the long term. You know, we know that optics, this dramatic amount of bandwidth, you know, moving from photons to electrons, crossing the optical electrical boundary, this was hard. Bringing light and bandwidth and low power right up to the package, and then scaling it for the industry to participate in it. Here we are, and we can now show this breakthrough. Our engineering team has done it, and they've built a detachable optical in-package connector coming directly into a glass substrate. Let's go live to our lab in Scotland and take a look to show it for the very first time. Russell.
Hi, Pat. Hi, Ria. Welcome to Innovation in Scotland. What we're gonna do today is. Yeah. Thanks, everybody. What we have behind us is a test and measurement system where we're gonna test that detachable connector. We have on the left-hand side a power meter. What you're gonna see is the power should flick up to the right-hand side. You'll see the needle in the power meter. On the right-hand side, the signal will go from red to green as we make the connection directly to the package with our detachable connector. What you can see here with John, we have the demo unit here. We have a silicon photonics package. We have optical fiber that runs up through the connector. It's a very small form factor detachable connector.
We're gonna measure the light that comes in and out of the silicon photonics chip with this setup. Okay? First time we've done this live, obviously. John's gonna make the connection now. Keep an eye on the screen. You'll be able to see the numbers as we make the connection. Here we go. We got power coming into directly the silicon photonics package. The green signal, that means we're measuring light all the way through the package. We have to show that it's also detachable. We're gonna detach it, reattach it again. I will be able to show that here now live. Okay, John. There we go. We've detached it. The signal's dropped off, reattaching it, and now the signal's jumping back up again. Okay? Great result, everybody.
Hey, Russell and team, you know, thank you so much. This is exactly how I dreamed of it, decades ago, and you guys are now delivering it. Thank you. Congratulations to our team in Scotland.
Thank you.
Absolutely. Thanks.
Can't believe that. I get to be part of this history-making moment.
Indeed, indeed. Ria, you know, thank you for joining me on this journey. You know, in 30 years, when you're on stage as the CEO, remember to continue to pass on the shared passion that we have for the world of technology and the enormous positive tech for good impact that it can have.
Oh, absolutely, Pat. I think on behalf of developers everywhere, we're ready to get these technologies running, working for our workflows, optimized, accelerated, and changing the world.
Maybe a memento for you.
All right.
Let's sign our poster here that we did.
Okay.
Grab your pen there, and you and I are gonna sign it, and we're gonna send you off here.
All right.
Right in this first superhero moment.
Okay. Gonna add a smiley.
All right. There we go.
Thank you.
Let's say thank you to Ria for joining us this morning.
Thank you all. Thank you, Pat.
Throughout our time together this morning, we've talked about our commitment to being open. Systems and software and hardware standards. Our collective potential as an industry is unleashed when we enable openness, choice, and trust. To close out, I wanna bring on stage a special guest, someone that I have personally admired for decades. No bigger champion of open collaborative ecosystems, the godfather of the open source movement, the creator of Linux kernel and of Git. Please join me in welcoming none other than Linus Torvalds.
Hi, Pat. Thank you.
You know, Linus, it is so great to have you at Intel Innovation. You know, we've both been around the industry a few years.
Yes.
You remember this?
I do remember that. I lost my own copy a few years ago, apparently.
Well, today we are fixing that.
Thank you.
I've signed this for you here.
Thank you.
When I wrote this book, it was the bible for the 80386, and this was really your beginning.
Literally, this was the thing I used on the project that wasn't Linux yet. Before I actually started doing an operating system, I was really just trying to learn about the new PC I'd bought. This was the CPU side of that journey, right? It took a while to learn the CPU and realize that what I'm actually writing is an operating system. This, in a sense, predates my Linux work. It was very integral to it.
Yeah. You know, maybe you could just take us back to 1991 when you were first getting started on the project. Tell us a little bit about that.
It's hard to even think back now because it was different for me. I came from a home PC kind of background and did not know the x86 architecture at all, but decided that this was the way into it and bought my first 386 and couldn't afford the commercial Unixes I wanted.
Ah.
And, uh-
You were cheap.
I was cheap.
Broke.
No. I prefer the word poor.
Poor. Okay.
All right.
Yeah.
I'd grown up with computers since I was about 11 and was used to just writing my own projects. The fact that I could not afford Unix myself, and I had this shiny new PC, and I had been exploring the CPU side of it. Because the 386 has the task switching hardware built in, it had been one of the things I'd been exploring. When you do task switching and do all the low-level, grotty paging details that
I expect.
This book literally goes through and explains, you kind of end up with a project that is the beginning of a kernel.
Yeah. Looks like an operating system.
Yeah.
Here we are today. You know, tell us what are the must-dos for a thriving open source ecosystem?
I think the must-do. One thing I really strive for is to make sure that people feel like they are part of. I don't want to say the community. I mean, it is part of the community, but they're part of making decisions and they really own this, the end result.
Invested in it.
They're invested in it. It's obviously why I chose the license I chose, making sure that I don't have any special rights. I really love open source. I love working in that community where it's about open discussion and really getting things done together. That's actually surprising to me even now, because I always, and I still feel that I'm not a people person. At the same time.
I've heard that criticism of you.
Yeah. No, I'm not the only one who thinks I'm not a people person. At the same time, it's been what has really motivated me for the last 30+ years.
Let's go forward 10, 20 years. What's your vision for open source in the future?
Yeah. No, this is not me. I'm not a visionary. I call myself a plodding engineer.
Plodding engineer.
I say that with pride. I think you need the people who have the big ideas, but you definitely in engineering, a lot of it is getting all the small details right and looking right in front of your feet instead of looking 20 years out.
Oh.
I look a couple of releases ahead, and I kind of have a notion of what's coming. My horizon is about six months away, not 20 years away.
Your impact, the industry that you've affected is nothing short of spectacular, Linus. You know, I mean, you and I, we're not finished yet. We got a lot of time in front of us yet. We'll be realizing the impact of your contributions for many, many years to come. It's with that in mind that today we are honored to give you and recognize you in our community as we come here together, right? You know, we would like to say that you are the first recipient of the first ever Intel Innovation Award.
Thank you, Pat. I don't know what to say. That is actually very nice and classy. We had a water bottle as a stand-in yesterday during the rehearsal.
This one's a little nicer. Yeah.
Which wasn't so nice and classy.
You know, while I was absolutely thrilled to have this conversation with Linus and I this morning, we recorded a long interview yesterday. We'll be posting that on our website. Truly, I couldn't think of a more honoring person for an open community than Linus Torvalds. Thank you so much for joining us today.
Thank you, Pat. Thank you.
You know, we've had an incredible morning together, but this is just the beginning. We're just getting started. On the next two days, labs, demos, show floor, cool, geeky things that we're gonna do. I encourage you to sign up for the developer cloud. You know, download our oneAPI Toolkit, you know, and engage with us, because our objective here is that developers, whether software or hardware, you see the future, and our job at Intel is to open that future up to you, working together on open frameworks that you can trust. I'm excited we've had the opportunity to come together to learn, grow, build, challenge, and help each other. Together, we've taken a peek into the future, one that we will create together. Thank you.
Day one keynote is in the books, and there is so much to unpack. Welcome to our post-show, Innovation 2022, live from San José. What a great keynote. Pat is on his way back to our blue carpet to debrief, so don't go anywhere. We saw a lot of the same themes from his keynote last year, but there was one fresh focus on this new almost post-pandemic era. Technology is increasingly central to every aspect of human existence. As we look ahead to the next decade, we will now see continued moves to digital for everything, the way we work, learn, connect, develop, and operate. All right. Let's bring Ravi back to get his reaction to this live keynote. Ravi, welcome back. What I thought was so cool was having this live audience there this year.
Yeah.
You know? That was an element we missed last year, being completely virtual with COVID.
Yeah, Madeline, it was amazing. You know, I gotta tell you, there were so many things that resonated. One, just the reaction to many of the things that Pat delivered. Two, the fact that Pat, you know, went after developer challenges, right? In a really, you know, phased approach, an organized approach, and then showing how, you know, Intel could, you know, enable solutions. That pop-up, you know, a comment that he made about his granddaughter, that was really-
Yeah.
inspiring for me.
Yeah. Thank you. When she puts her USB in. That was precious. You know, a part of this live show element is the audience reactions, and I was really impressed by the excitement we felt when he said that, when he talked about Andy Grove and the speed of innovation and how all of that is still alive and well.
Hey, hey, guess who's standing next to me?
Oh.
Pat.
Oh, hey.
That was amazing. I could feel the excitement in there. You know, you tell me, again, coming back, looking at all the developers, you know, the folks, the reaction, how do you feel after that keynote?
Oh, I feel great. Right. You know, to me, it really is this coming back to this community that is looking for us to be engaging, to help to create these open, technologies, for tomorrow. We had some cool demos to show off and some real progress since last year, and, just setting up for the second half tomorrow.
Yeah. Very, very cool demos out there, by the way. You know, last year you said you were bringing back the geek, right? As our chief geek, h ow have the developer community rallied around the things that we're doing, and how have they responded to all of the things that we've been doing over the last 12 months and here recently?
Well, you know, I'd say, you know, at first they're sort of like, "Huh, are you really back?" Right? You know, waiting to see. We had a lot of work to do to get the underpinnings in place. You know, we said oneAPI and, you know, the oneAPI is there, the, you know, regular release is there. So that's happening now. We said DevCloud. Okay, well, that's happening now. We said we're gonna have an open AI stack. Well, we didn't have a lot of pieces in place to make that possible. So now it's all sort of coming, you know, together, and they're responding super well. You know, things like the DevCloud, you know, we have a queue of people ready to get on board and start getting access to our latest and greatest.
That's amazing. Hey, I've got to tell you, Madeline and I were just speaking, that comment you made about Pop Pop and your granddaughter, I mean, really amazing, and I hope she's watching. You had Ria and Linus on stage. What was your inspiration for having both of them there with you front and center this year?
Well, you know, first, to me, it always has been this amazing thing. I mean, Andy sort of reached down and grabbed me when I was a 22-year-old engineer, and I've always had that view of Intel, you know, that, hey, I don't care if you're, you know, 22 or 62, right? The question is, are you a geek and are you gonna be able to do some amazing things? Our company just enables that. Having somebody which is like, I say I'm an underachiever, you know, compared to Ria, right? What she's accomplishing. I want to be the place that she can really, you know, fulfill, you know, her full potential as a leader. When we flip to the other side with Linus, you know, truly is somebody that, you know, I've admired for decades.
If you're going to talk about open, who else could you have that's more representative? You know, Linus doesn't do these things very often. He said maybe once or twice a year. You know, it's a very unusual thing. Tracing our roots back to the earliest days of the 386, we have sort of this unique bond as well.
Yeah. That's amazing. I mean, transparency, openness, innovation. Pat, thank you very much. We'll see you later on in the show.
Thank you, Ravi.
Yeah. Hey, Madeline, back to you.
How exciting and cool that we had Pat live to debrief. What a cool concept of Andy Grove reaching to Pat and now Pat maybe reaching to Ria, maybe our future CEO. Who knows?
Yeah, it's an interesting commentary, definitely.
Awesome. Well, I know you need to get back here for the rest of the show. We'll see you back here in a minute. On to the highlights. The Intel Developer Cloud has expanded to give developers and partners early access to new and future Intel technologies. This gives developers a head start on product pre-qualification, anywhere from a few months to up to a full year before the product is available. The beta trial is now open to pre-qualified Intel customers and developers. That means getting your hands on the Intel 4th Gen Xeon Scalable processors, Intel Xeon D processors, the Habana Gaudi2 training processor, the Intel Ponte Vecchio Data Center GPU, and the Intel Data Center GPU Flex Series. Intel is delivering the tools developers need to boost productivity and stay one step ahead. Intel also has a long history collaborating on standards bodies worldwide.
The company is leading the launch of the Universal Chiplet Interconnect Express, or UCIe, to create an open chiplet ecosystem. They're delivering a standard to allow more flexibility in chiplet construction, so that chiplets or components from different vendors can integrate into advanced packaging technologies. That means you either get chiplets or power supply components from Intel or one of its competitors, and Intel, with its leading 3D packaging technology, will assemble them into the marketplace. I think of this as a prime example of coopetition. Intel also revealed major milestones across its line of data center GPUs in addition to pricing and availability for the first Arc GPUs for gaming. The Arc A770 is Intel's top-end gaming desktop GPU. You heard the crowd go really wild for this one. The A770 will be available in retail starting October 12 at only $329.
That strikes a balance between pricing and performance. The Intel Data Center GPU Flex Series will now run popular industry AI and deep learning frameworks like OpenVINO, TensorFlow, and PyTorch. Through the new Intel On Demand activation model, customers can turn on additional accelerators beyond the base configuration for greater flexibility and choice when needed. Let's not forget the big hardware news today. I think we heard the most cheers on this one. Intel unveiled the 13th Gen Core K Series processor with an exciting gaming demo. Earlier, I caught up with Executive Vice President and General Manager of the Client Computing Group, Michelle Johnston Holthaus, for an exclusive look at what these processors can do. Let's take a look. Michelle, thank you so much for joining us in person here at the showcase. We're excited to have you.
Thrilled to be here.
All right. 13th Gen, it is official. What do developers need to know?
Yes, I'm so excited. The 13th Intel Gen processors are here, and wow, do they pack a punch. Whether you're a modern gamer that's looking for that amazing single-threaded performance, we give you 15% more generation on generation. If you're a creator looking for more E-cores to be able to do that multi-workload experience, this product has it. If you're an enthusiast and you wanna overclock, whether you're a beginner or you're well-versed in overclocking, this platform has it all. It comes in a variety of price points from i5 through i9.
Madeline, I'm so excited to get this product in the hands of our consumers and customers.
We were using the phrase earlier that I love, mega tasking, when it comes to this product. Think I'm gonna be using that from now on.
Yes. Everyone should be using that, and everyone should be going out and getting this CPU in October.
Tell us about Unison?
Yeah, excited about Unison. Think about multi-device. When you have your PC and your phone, and you're switching between two, they're not communicating and sharing information back and forth.
Right
Like we would all like. Unison fixed that. Whether you're on iOS, Android, you're gonna get a seamless, ubiquitous experience across the platform. Launching later this year with Acer, Lenovo, and HP, just to name a few. I'm very excited about this new product and bringing a better platform experience to all of our customers.
Super exciting. Let's talk about something else that's exciting. Last we talked at Vision, you were only a few months into this new role leading CCG. Now you're about six months in. How's it going?
It's going great. It's so fun to see the amount of innovation, the excitement from the engineers in CCG, and they're just complete passion to deliver products that our customers need. We're seeing the first one come to fruition, and I can't wait for many more.
There's a lot to celebrate.
There is a lot to celebrate. Yes, absolutely.
Yeah. I'll say that I miss you over in SMG. I think we all do, but we're really proud of you.
Aw
leading CCG.
Thank you. Thrilled to be here.
Thanks for being with us.
Yes.
There's a lot more to come. Stay with us.
Thanks for joining.
We are just getting started here at Innovation Live. Coming up, one of the stars of today's keynote, the brilliant Ria Cheruvu, will join us on set. Greg Lavender will preview his day two keynote, and we will meet the winners of the AI Hackathon. Join the conversation at #IntelON, and we'll be right back.
There's an innovator in all of us, and your brightest ideas can come in an instant. We'll stop at nothing to provide intelligent technology that optimizes user experience, strengthens security, and maximizes your storage, so you can bring your greatest ideas to life. [Foreign language].
The technology inside, the technology that creates worlds. How wonderful is that? Intel.
What if you were a major transit system with billions of passengers taking millions of trips every year? You weren't about to let any cyber attacks slow you down, so you partner with IBM.
This is driven by what I like to call the superpowers, these foundational technologies that are profoundly shaping how we experience the world, and it creates this bridge from the analog, from the physical to the digital world. Each one of these five powerfully, individually, you know, just changing. As you think about it, more sensing means more compute. More compute means more connectivity. More connectivity, more infrastructure. More data over that infrastructure, more AI, which leads to more use cases. Right? They're reinforcing, accelerating, empowering each other to unlock even more powerful possibilities in the future.
All right, welcome back. I now officially have Ravi on set.
Hey, great to be back with you.
Welcome. Good to see you. Backstage, you had some great interviews back there.
Oh, it was amazing. Amazing.
Wow. Let's respond to this really quickly. We have another superpower, sensing. How incredible is that? I thought it was cool in the keynote how he talked about sensing builds on all of the other superpowers. All of the other superpowers will be required to now have this fifth superpower of sensing, of Web 3.0, of the metaverse, of feeling.
Right. Yeah, definitely taking it deeper and broader. Everything we've been hearing this morning in terms of AI and Ria, sensing is just the next step. Yeah, definitely an evolution to the superpowers, and it's great to hear it.
Absolutely. When you think about how this integrates with Intel's commitment for tech for good and people with disabilities and accessibility on so many levels. You know, Pat talking about his own hearing aid, I thought that was a really interesting application-
Yeah. Yeah
of this superpower and where we see the future of technology going.
Right. Yeah, it's enabling so many things, and when we think about Intel, we think about technology, we think a lot about gaming. You know, when you start to put the human element in terms of medical and those things, it's just really amazing to hear and see all the things that we're doing in that space enabling those capabilities.
Yeah. Another cool moment in the keynote was when Pat gave the first ever Lifetime Achievement Award to Linus Torvalds. It was so interesting because, you know, Linus and Pat go way back, way back. Linus has also had some not so positive things to say about Intel.
Yeah
You know, in the recent years.
Yeah.
So-
No, I thought it was a bold move. I mean, we talk about transparency, open ecosystems and, you know, having Linus there and, you know, on stage with Pat, it was just, you know, another testament to us listening to the community and-
Yeah
showing up with the community.
Absolutely. I think that's so well said. It's such a testament to us being a part of this community, integrated into this community, and it being about the community. For developers, by developers.
Right.
All right. Now we have a very special guest. I am so excited to welcome the brilliant Ria straight off the keynote stage. Welcome. How are you feeling after that keynote?
Thank you.
That was incredible.
I'm feeling great. Definitely my first time being on a stage that big.
Yeah.
Yeah, I'm very happy receiving lots of awesome notes from colleagues and family.
It won't be your last time. Now that you've gotten the bug, right? You're gonna be up on that stage every year.
Oh, definitely. I mean, that's my plan.
Well, I think Pat maybe just said you were gonna be CEO.
Me? I have bold ambitions, but I know I have a lot to do to get there and to learn as well, so I'm just excited to take it all in.
Absolutely. Well, I know you had a chance to catch up with Ravi earlier, but Ravi, did you have another question you wanted to ask?
Well, one, you're a natural, and you have amazing presence out there.
Thank you.
What you're really amazing at doing is bringing these things that are really complex, and you know, articulating in a way that, you know, all of us, even us non-developers can understand.
Right.
really enjoyed hearing you up there.
Awesome.
Spending time with you also.
Thank you so much. Communication is definitely my passion. I mean, different types of forms and mediums, I'm all for it, so.
Tell us, what was your inspiration to get into, you know, responsible AI, ethical AI, when you could have done anything in the world?
Absolutely. It was a kind of interesting transition because I started off with, you know, just like the typical kind of deep learning and hardware acceleration types of items. I jumped over then to the labs, you know, like research collaboration on computer physics and reinforcement learning. Eventually I started to think about, you know, the different trajectories and the impact of the AI systems at large. That brought me to responsible AI and thinking, "Okay, what can we do?" At that time, Lama Nachman and team were like, you know, forming and creating and also, you know, establishing a lot of the key parts of responsible AI at Intel. I was able to join that team and help contribute to the effort from a technical side.
Well, it's so interesting because I feel like a part of what drives you is this human element. When you had your one-on-one with Pat in December, you shared with us how his advice to you was understanding how people think and the psychology. Could you tell us a little bit more about your conversation with Pat and his advice to you as the future?
Absolutely. Yeah. I think the core part that I really pulled out of that is diversification. Also, you know, something that's a little more subtle, but I'm just pulling from Pat directly when he's on stage and interacting with him, is that enthusiasm for technology, for developers, and for the folks that are watching and getting empowered by these technologies. I think that type of perspective to be able to fan out and take on new challenges, I'm definitely inspired by that. Again, you know, I see a lot that needs to be done, and I'm ready to go and do it, so.
Amazing. Yeah. You know, you interact with a lot of people. You're getting a lot of this input. You know, let me ask you, people's reactions to you must always be, you know, just stunned, amazed, right? I'm amazed, right?
Yeah.
I should be. You know, how do you react to that? Does that ever get old? You know, how do you react to this constant, "You're so young. It's amazing, da, da, da." Give us a little insight into that.
Oh, it's absolutely fantastic. Never get tired of it because it's like the power that's keeping me going. It's the right word is blessing, definitely. It's like the collective kind of energy, enthusiasm that I'm just, you know, consuming and then bringing on to my presence. I'm always for it. If I need to do something right, I'm for that feedback. But if I did do something right, then I wanna keep it happening. That's what I need my team, my family, my colleagues. One Intel in like a one industry. I'm really for that concept. Yeah.
Especially your mom, right?
Oh my gosh, yes. She is the best. She's my behind the scenes everything, so like I love her. She's always there for me. She was like front row in the keynote as well. I haven't. Yeah.
I know she must be proud. We're all proud, and we really appreciate you stopping by here first.
Oh, thanks. Definitely.
After the keynote. More to come from you, including a book of poetry, I hear.
Oh yeah, absolutely. Yeah. More books around like machine learning for young minds and all sorts of stuff, and poetry as well on the artistic side. Yeah. Awesome.
More to come. Thanks again. Thanks for being here.
Thank you.
Shifting gears, this might have been one of the most groundbreaking demos that I've seen on stage. From their lab outside of Edinburgh, Scotland, Intel tested their detachable optical in-packaging connector for the very first time. We saw Pat predict this a few decades ago, and then it was incredible to see this live. Just phenomenal.
Yeah.
All right, let's look at the tape.
Engineering team has done it, and they've built a detachable optical in-package connector coming directly into a glass substrate. Let's go live to our lab in Scotland and take a look to show it for the very first time. Russell?
Hi, Pat. Hi, Ria. Welcome to Innovation in Scotland. What we're gonna do today is. Thanks, everybody. What we have behind us is a test and measurement system where we're gonna test that detachable connector. We have on the left-hand side a power meter. What you're gonna see is the power should flick up to the right-hand side. You'll see the needle in the power meter. On the right-hand side, the signal will go from red to green as we make a connection directly to the package with our detachable connector. What you can see here with John, we have the demo unit here. We have a silicon photonics package. We have optical fiber that runs up through the connector. It's a very small form factor detachable connector.
We're gonna measure the light that comes in and out of the silicon photonics chip with this setup. Okay? First time we've done this live, obviously. John's gonna make the connection now. Keep an eye on the screen. You'll be able to see the numbers as we make the connection. Here we go. We got power coming into directly the silicon photonics package. The green signal, that means we're measuring light all the way through the package. We have to show that it's also detachable. We're gonna detach it, reattach it again. I will be able to show that here now live. Okay, John. There we go. We've detached it. The signal's dropped off. Reattaching it, and now the signal's jumping back up again. Okay. Great result, everybody.
Hey, Russ ell and team, you know, thank you so much. This is exactly how I dreamed of it, decades ago, and you guys are now delivering it. Thank you. Congratulations to your team in Scotland.
Thank you.
You know, Ravi, the beauty of live TV is when they disconnected it and reconnected it. It's like fingers crossed this is gonna work, and the crowd went wild.
Right. That was an amazing moment.
It was.
Amazing.
It was. The beauty of live demos. Coming up, we will be joined by tomorrow's keynote speaker, Greg Lavender. Stick with us. We'll be right back.
71% of enterprises are repatriating cloud workloads to improve cost and control. Learn more at www.supermicro.com.
Bum, bum, bum.
The technology inside, the technology that inspires. How wonderful is that? Intel.
Because we've been this way from the start. Because the solution for one can be harnessed by others. Because clouds that compete can still connect. Because innovation needs collaboration. Because a revolutionary approach to software can transcend enterprise IT. Red Hat. Our code is open.
This moment really showcased the magic of technology. JS Choi, CEO of Samsung Display, joined the keynote to showcase the first 17-in slidable display globally for the first time. Instead of folding the screen, it slides or rolls out to give the benefit of a larger screen, but also the convenience of portability. I can't believe. Who did you have to bribe to get one of these on set?
I love this. I think I'm gonna take this home.
Well, I think Ria already claimed it a couple of times on the keynote. I heard that.
Look at this. Look at that.
Let's see it in action. Man, is it easy to pull, too?
Yeah, very easy.
Seamless, as seamless as it looks?
Yeah. Could you imagine just having this on your desk and needing to make space?
Well, I loved what Craig said about, you know, no TSA.
Right.
No back problems.
Right. I mean, that's amazing.
For those of us always-
Yeah, look at that.
You know, on the road, yeah, that's really, really cool.
Yeah.
Are you gonna keep? Are you hiding that somewhere?
I don't know. I'm trying to keep it from you.
To get it. Yeah.
Did you wanna hold it or
I do. Yeah, I do.
Okay. Okay.
Okay, I'm gonna try not to break it. All right, my turn. Oh, yeah, y'all, it really is actually that easy.
Isn't that easy?
Yeah, it is.
Yeah, amazing technology.
Here we go. Yeah, I'm being told from backstage.
Where's your purse?
Maybe I can put it in my purse that's back here behind the stage. Either way.
Yeah.
I think this might come home. All right, y'all, we're gonna keep this innovation party going. Join me on Friday along with special guests Arun Gupta, Bill Pearson, and Jen Huffstetler on Twitter Spaces for hashtag Processing Out Loud. We'll go live at 10:00 A.M. Pacific. Submit your questions in advance and meet us there. Earlier today, I also caught up with Greg Lavender, and he introduced me to some very special guests. Let's take a look. I am so excited to have our Chief Technology Officer, Greg Lavender, here joining me on set. Thank you for being here with us.
Glad to be here, and excited to be here for Intel Innovation. It's our second Intel Innovation.
Absolutely. Well, we're super excited for day two keynote. This is your baby. What can you tell us about it?
Well, I don't wanna preview everything, but let's just say I'm gonna have some guests with me online, you know, online and in person. We'll be talking about, you know, some cool technology, right? Giving you the preview of our confidential computing technology known as Project Amber. We have of course some AI, ML, you know, announcements to make, along with some customers who've actually doing things and using things. I'm really excited by some of the things we're gonna talk about that for developers by developers and really kinda make it something that people will be surprised by and that they're gonna really enjoy participating with Intel to get real value for what they're doing.
I think we're hammering home, right? The developers wanna write code, make money, do good, right?
You wanna do it quickly.
Do it quickly.
You wanna do it quickly, and you wanna be using the latest hardware that you can get.
We can help them with that, right?
Absolutely. Stay tuned.
Stay tuned. All right, speaking of guests, you brought a few guests with you today.
We got to our hackathon. We did an AI hackathon. We had a little over 100 people participating here in San José. I'm happy to introduce Ankur and Raju, who basically won the hackathon as a team. They're grad students at San José State University studying data science. You know, appreciate your participation. Obviously, I think it went for 12 hours or something like that.
Yeah. Almost. Almost, yeah.
Almost 12 hours, so that's quite a hackathon. I mean, I'm used to 24-hour hackathons, but that's called just coding. As a special prize I want to offer, they're both gonna receive one of our Intel 12th Gen Intel NUCs, which is our latest. We call it Dragon Canyon. It's our latest Alder Lake processors. I have one of these myself. I sort of, when I first got it, I just basically disassembled the whole thing, pulled it all apart to see how it was put together. It's an amazing piece of engineering, both thermally and mechanically, in terms of packing so much compute power and capability into a small form factor compute device. It's the dream compute machine for Intel developers developing on our latest processors.
Of course, you can have your selection of GPU that you want to add to it. I think you'll both get one, you'll both enjoy it, and you'll be the envy of your friends.
I know y'all are so excited. You were hoping that your microphones would stretch to get your hands on it sooner.
I know.
You can touch it in a little bit.
Yeah, congratulations. Congratulations.
Absolutely.
We'd love to ask you a couple of questions.
Sure.
Ankur, you said you were surprised by the processing speed.
Mm-hmm. Mm-hmm.
How was the experience?
Before I answer that, I'd like to start by thanking everyone, and I'm really glad to be here. I honestly didn't expect to win the grand prize, but I had my eyes on one of those things. It turned out good.
Motivating.
Yes. Answering your question, I think, like, that kind of speeds, like I have only seen in GPUs, so I was quite taken, like, it was kind of jaw-dropping for me. For a second, I thought they have some other thing, in the back end, but in the slides they have shown they are using only Intel chips, so that was kind of surprising. Just to give you some context, we processed like 1,300 images in just. Each epoch was, like, taking one second. We used a ConvNeXt tiny model, so it is quite a big model for a CPU, and still the epochs time was fairly less. I was actually taken aback by the, like, accelerations that, Intel Extension for PyTorch provides. Great job on that.
Awesome. Thank you.
Great. Thank you.
You said it was especially interesting for you and exciting because as college students-
Mm-hmm
you don't get access to these kind of, you know, processing technologies, capabilities.
Mm-hmm
toolkits.
Mm-hmm.
Yeah. It's actually a hidden gem. Intel oneAPI is like. Because probably, people who are already there in the industry probably might know about this, such acceleration and probably, we can do things on CPU itself. We have this quote in our machine learning community that most of the 80% of the work can be done on by just the simple ML models. No need for the deep learning stuff. ML models, we have oneAPI for even for the ML models for scikit-learn and also the gradient boosting. I actually got luckily, because of my professor, I got actually introduced to this Intel acceleration kit in my previous semester. That is how I got the initial touch.
I mean, documentation provided by Intel is, like, very good. It's one thing that is best about this acceleration part is, like, it takes away all the. It doesn't break your code. It's just one or two lines of code which you add, and it just does the magic behind the scenes, and you do not need to focus on integrating it into your code without breaking your code. The documentation is also, like, it explains very neatly. Yeah, it's. I truly feel the Intel acceleration is kind of a hidden gem, at least for the academia side.
Mm-hmm.
Because in academia, we do not generally think of in a deployment perspective. We do not think about speed. It's just like we have to build models and we have to leave those.
Yeah.
Yeah.
Sounds like a good market for Intel, right?
Well, it's gratifying to actually hear you say that because, again, we've invested a tremendous amount of engineering effort into the open source ecosystems.
Mm-hmm.
PyTorch, TensorFlow, in fact, TensorFlow 2.9 out of the box has our oneDNN, deep neural network library acceleration, right, for our processors. You just take it, you don't have to change anything. You just basically get the benefit of it running on Intel processors.
Okay.
We have, you know, XGBoost, you know, scikit-learn. We invest in all of these PaddlePaddle. You know, we invest lots of optimization and for free put that code out there to the open ecosystem so that students like you and your friends can basically make a lot of progress using our, you know, commodity processors, right? You'll get to hear more about what we're delivering through the cloud services and our next generation Xeon processor, which is gonna have even more push and more hardware accelerators to take advantage of. It'll all be accessible, you know, shortly through Intel Developer Cloud as well as through our cloud service provider partners and OEMs and ODMs.
Really excited by all the great hardware acceleration we're delivering to the AI ML community and a full service AI everywhere stack to let you run as efficiently and successfully as possible.
Really exciting news. Really exciting award. Congratulations again. Thank you, Ankur and Raju, for being here today.
Thank you.
Thank you, Greg, for being here today.
Absolutely.
More to come at Intel Innovation. All right. Coming up tomorrow, 8:00 A.M. Pacific, same time, same place, Innovation Live with Madeline and Ravi. That's us. At 8:30 A.M., we'll kick off with the Greg Lavender keynote. At 9:30 A.M., the Andrew Ng luminaries, and then at 10:15 A.M., we're back here with Innovation Live with Madeline and Ravi. All right. Thanks, Ravi, for being here today. I'm so excited to have you here as a co-host.
Thank you for having me. It's been wonderful, and I look forward to tomorrow.
Successful day one. See you back here for day two tomorrow. That's it for us.