At this year’s CES, Dr. Lisa Su, CEO of AMD, gave a pre-show keynote speech, and she brought an endless stream of announcements and guests. The theme of the keynote is high performance and adaptive computing to help solve problems. Guests on stage included representatives from HP, Intuitive Surgical, Lenovo, Magic Leap and Microsoft. It also includes a former NASA astronaut for the bonus.
Dr. Su made seven key product announcements ranging from energy-efficient PCs to extreme processors designed for supercomputers. With all the announcements and guests, it’s no wonder the keynote took a lot longer than planned. It seemed like every group at AMD was involved in the keynote.
Despite the current decline in PC sales, most of AMD’s announcements have been focused on PCs — both desktops and laptops. Of particular interest is AMD’s decision to add dedicated AI cores to its new Ryzen 7040-series notebook processors. AMD Ryzen 7040 is designed for ultra-thin notebooks with a TDP (Thermal Design Power) of 15W to 45W and is fabricated on TSMC’s advanced 4nm process. The Ryzen 7040 features AMD’s most advanced Zen 4 CPU cores paired with the latest RDNA3 GPU architecture.
The Ryzen 7040’s embedded AI technology was developed by Xilinx, which AMD acquired last year. But even before the acquisition, AMD had partnered with Xilinx to license the technology. Xilinx has developed an AI engine for its Versal ACAP adaptive computing products. Details of what AMD is now calling the Ryzen AI engine are still to be revealed, but the Versal AI engine is a two-dimensional array of VLIW vector-vector and matrix-matrix compute engines. While AI workloads can run on AMD’s CPUs and integrated GPUs, AMD claims it’s more energy-efficient to run these workloads on a dedicated AI engine, which is why the company is pioneering it in notebook processors. Panos Panay, Microsoft’s executive vice president and chief product officer, gave a big nod to Ryzen AI and how Microsoft plans to use it initially for Teams and other features in the future. We encourage AMD to provide access to Ryzen AI through open toolchains and APIs in addition to Microsoft’s Studio tools.
It’s great to see AMD really embracing the pervasiveness of AI in computing. AMD has GPUs and FPGAs/ACAPs for AI acceleration, and is now incorporating AI technology into mainstream notebook processors. Intel will also integrate an artificial intelligence accelerator into its Meteor Lake processors later this year.
AMD also brought its RDNA 3 architecture to mobile discrete GPUs with the release of the Radeon 7000M series. These new mobile GPUs can power 1080p gaming at maximum game settings and support graphics-intensive content creation applications. When you combine AMD Radeon RX 7000M Series graphics cards with Ryzen 7000 Series processors, AMD SmartShift RSR technology intelligently distributes image rendering, upscaling and presentation demands across APU and GPU resources to optimize performance. The feature is expected to be available in the first half of 2023.
Desktops are also getting processor upgrades. The new Ryzen 7000 series X3D desktop processors are equipped with up to 144MB cache and support up to 16 cores and 32 threads. The new processor is a bit of an oddity, as only one of the two CPU cluster chiplets in the package gets stacked memory, giving the CPU an asymmetric bias. For example, the Ryzen 9 7950X3D has an 8-core CPU chiplet without 3D cache that can hit 5.7GHz when boosted, but another 8-core chiplet with stacked 3D cache in the processor has more cache memory, but boosts the clock lower. It will be interesting to see how OS thread managers and application developers handle this asymmetrical configuration.
Other PC-related announcements include the high-power Ryzen 7045HX notebook processor designed for larger laptops, aimed at gamers and creative professionals. The processor uses a chiplet design from the desktop segment and offers up to 16 cores.
It also announced in a low-key manner that the AMD Ryzen 7000 series desktop processors will be available soon. Based on the Zen 4 architecture and featuring a 65W TDP, the new Ryzen 7000 series processors are optimized for efficiency and performance. There will also be some cost-optimized AM5 motherboards before spring.
AI in the Data Center
Three announcements about how AMD is working to advance AI for data center applications: the AMD Vitis Medical Imaging Library, the Alveo V70 AI Accelerator, and the Instinct MI300 High Performance Computing Accelerator.
AMD announced its Vitis medical imaging library to bring high-quality medical imaging products to market faster by reducing development time. These software libraries run on AMD Versal SoC devices with AI Engines, enabling high-quality, low-latency imaging for medical applications.
AMD also demonstrated its XDNA adaptive AI architecture in a discrete PCIe add-in card with Alveo V70. Based on AMD XDNA with AI Engine architecture, the Alveo V70 extends ubiquitous AI from the edge to the cloud.
Xilinx FPGAs are currently used in the Perseverance Mars Rover, so AMD hosted an interview with former NASA astronaut Dr. Cady Coleman. Dr. Coleman is advancing NASA’s agenda for human space travel to the moon and eventually to Mars through the Artemis program. But the broader message is about promoting STEM education and inspiring more people, including women, to pursue careers in science.
Dr. Su concluded her keynote with a preview of the company’s next-generation GPU for high-performance computing — the MI300 — which consists of 13 chiplets and uses 2.5D and 3D packaging technologies. It has a total of 146 billion transistors. The MI300 not only offers GPU cores, but also includes three Zen 4 CPU chipsets. The purpose of encapsulating CPU cores in a GPU cluster is to be able to run tasks closer to the GPUs and eliminate transactional latency with the main host CPU. 128GB of HBM (High Bandwidth Memory) local memory in the package will support CPU/GPU clusters. The design is similar to Nvidia’s Grace-Hopper HPC processor, which combines Grace’s ARM CPU with a Hopper GPU in the same package. The MI300 should ship in 2H23.
Numerous presentations during Dr. Su’s keynote address demonstrate that AMD will continue to deliver leading products in PCs and high-performance computing. Despite the current tech market downturn, AMD continues to invest in new cutting-edge products and is accelerating its AI roadmap.