AMD keynote at CES 2023 in 10 minutes

AMD keynote at CES 2023 in 10 minutes

Video Transcript

- And I'm really excited to be joined by some of our closest partners to discuss what we're doing across AI, hybrid work, gaming, health care, aerospace, and sustainable computing. So let me start today with a brand new accelerator that we're bringing into the market.

AMD XDNA is a highly-configurable AI accelerator that originated from our acquisition of Xilinx. It actually allows us to scale from PCs to intelligent endpoints to edge devices and even into the cloud. The first use of this architecture in AMD products will be in our Ryzen mobile CPUs for notebooks. So today I'm very proud to announce the all new Ryzen 7040 series.



And it's the industry's first mobile x86 processor to integrate a dedicated on-chip AI engine which we call Ryzen AI. It features up to eight Zen 4 cores, our latest RDNA three graphics, and our new Ryzen AI engine that runs four dedicated AI streams simultaneously, delivering up to 12 trillion AI operations per second. Here is the all new 7040.


This chip uses four nanometer process technology and it has more than 25 billion transistors, almost twice as many as our Ryzen 6000 generation. The Ryzen 7040 series will enable more than 30 hours of battery life, which is just simply fantastic. The first rise in 7040 notebooks will be available in March, and we will have more than 250 ultra thin gaming and commercial notebook designs spanning our entire Ryzen 7000 series portfolio on track to launch this year.

- But now AMD is also at the forefront of AI technology with Ryzen, the 7040 series alongside Windows 11, it is our next step in this journey together.

- Today we announced the Dragonfly Pro product. By working together, we have been able to deliver 40% higher passcodes score than the M1 that was the leading product until now. You will be seeing a full portfolio of solutions in the spring that will leverage all these products.

- The Ryzen 7040HX series is our first mobile CPU based on a chiplet design. It's for gamers and creators who want the highest possible performance in a notebook. It combines up to 16 of our high-performance Zen 4 cores with boost speeds up to 5.4 gigahertz and up to 80 megabytes of on-chip memory in 5 nanometer process technology.

We're now launching our first products in laptops with our new Radeon RX 7000 mobile GPUs. And what we're doing here, the first chip in the series, the RX 7000 series for gaming and creator laptops is the new Radeon 7600MXT.

And what we have here is 32 RDNA three compute units, 8 gigabytes of GDDR6 memory, and a configurable power that allows us to adjust for the best balance of performance and battery life across a wide range of system designs.

You're going to see these guys in the market very, very soon. So we'll have Ryzen 7945 and Radeon 7600MXT laptops starting in February, and you'll see even more gaming notebooks coming later this year, including some new AMD advantage notebooks.

- And so tomorrow we'll be unveiling our first Lenovo Legion Pro Series laptops, and they integrate the latest generation of AMD Ryzen 7045 series and processors that you just launched, Lisa, to take gaming to the next level. And the new Ryzen 7045HX series bring fantastic best in class processors to gaming laptops.

- And tonight, I'm very excited to announce that we are now bringing 3D V-cache technology to our Ryzen 7000 processors.


The Ryzen 7 7800X3D is 8 cores, 16 threads, up to 5 gigahertz frequency, and 104 megabytes of cache, which more than doubles the cache compared to the Ryzen 7700X nonX3D version. And these larger caches are important, especially in gaming.

But guys, as great as the 7800X3D is, I've had a lot of our fans asking me for even higher-end options. So for tonight, I'm very happy to announce that we're also bringing 3D V-cache technology to 12 and 16 core Ryzen 7000 processors.


We've been working hard on this. The Ryzen 9 7950X3D is our first 16 core Ryzen processor with V-cache technology and our fastest 3D stack chip ever. It features 16 high-performance CPU cores, boost speeds up to 5.7 gigahertz, and a huge 144 megabyte cache.

So I'm very excited to announce that we will be bundling the game with our select Ryzen 7000 series CPU starting later this month because we're going to have an incredible portfolio for you with Ryzen desktop processors.

The new 7800, 7900, and 7950X3D parts will launch in February, and we are also launching new low-power 65 Watt Ryzen 7000 processors and an expanded portfolio of entry level M5 motherboards that will make Ryzen the best CPU at every price point in the desktop market.

We are very proud to be number one in adaptive computing, with 10 of the top auto manufacturers, six of the top seven 5G wireless equipment manufacturers, and more than 6,000 different customers across health care, aerospace, industrial, and other embedded markets that are using our products.

- Having a multi-generational, scalable, and reusable compute platform helps intuitive integrate the same parts across multiple product platforms. We've worked very closely together to integrate your adaptive technology. Handling multiple real time functions concurrently with real time compute and with low latency is key.

- And tonight, we are announcing new vitus medical imaging libraries that will make it even easier for medical device manufacturers to offer high resolution imaging at 1,000 frames per second or more.

- It was a game changer for us.

KRIS DICKIE: The AMD Zynq really helps solve our technical challenges by creating a platform where we can add software, automation, and artificial intelligence to bring all those controls and automations onto a single platform.

- Because of AMD's leadership in Zen cores and Radeon graphics, we partnered with your team to define a custom processor for Magic Leap 2. And we also developed a custom computer vision and AI processing engine that's been a key component to how Magic Leap 2 maps and then understands that physical world. And our collaboration with AMD has resulted in the industry's most advanced, augmented reality platform for enterprise.

I'm excited to announce that we are now what's called 60601 certified, and that means we're approved to take into the operating room, the surgeon can wear it in, and we're currently the only AI device with this certification in place. In fact, the Senti AI solution is now under that standard review by the FDA, and we hope to see their software on our platform a little bit later this year in the market.

- Perseverance Mars Rover, adaptive technologies are used to help Perseverance, or Percy, as many call her, to navigate on the surface of Mars using object recognition and terrain avoidance. So they used adaptive technology for systems like engine controllers for displays, for command systems. And the Orion capsule had more than 5,000 unique sensors. And that was just on the mannequins inside of the capsule.

- 4th Gen EPYC is truly the world's best data center CPU with more than 300 world records. It also delivers leadership energy efficiency that translates into the best total cost of ownership.

Today we're going to preview AMD's first inference accelerator. This is the AMD Alveo V70. It's built with AMD XDNA technology, and that's the same AI engine processor that we talked about with Ryzen AI. But it's scaled up for use and servers to deliver 400 million AI operations per second. That's 400 trillion AI operations per second.

We designed Alveo V70 to accelerate multiple AI models, including video analytics, customer recommendation engines while delivering just great compute efficiency in a small form factor. And I'm very pleased to announce that we're going to take pre-orders for the V70 cadence today with availability this spring.

Our Instinct MI300 is the first chip that brings together a data center CPU, GPU, and memory into a single, integrated design. What this allows us to do is share system resources for the memory and IO, and it results in a significant increase in performance and efficiency, as well as it's much easier to program.


I am super proud to show you MI300 for the very, very first time. This is actually the most complex chip we've ever built. It has nine 5 nanometer chiplets that are 3D stacked on top of four 6 nanometer triplets with significant HBM surrounding it, and it has more than 146 billion transistors.

So I'm very happy to say MI300 is currently in the labs and we'll be sampling to customers shortly, and you can expect that we'll bring MI300 to market in the second half of this year for both HPC and AI solutions.


Thank you for being such a great audience here in Las Vegas, and thank you to everyone watching online. Have a great night.