AMD launches new Epyc server CPUs integrated with MI accelerators, completing full-stack AI platform strategy.
AMD announced its next-generation Epyc server processor on October 10, built on the new Zen 5 core architecture with up to 192 cores designed for cloud, enterprise, and artificial intelligence workloads. The fifth-generation AMD Epyc 9005 series, formerly code-named Turin, can function as a standalone chip for general-purpose workloads or AI inferencing, or be paired with AI accelerators like AMD's Instinct Series GPUs for larger AI models and applications. At its Advancing AI event in San Francisco, AMD also announced new GPUs and data center networking solutions, including a new DPU and NIC, while reiterating its plan to release a new GPU every year, starting with the AMD Instinct MI325X accelerator in the fourth quarter of this year.
The new Epyc processor delivers up to 17% better instructions per clock (IPC) for enterprise and cloud workloads and up to 37% higher IPC for AI and high-performance computing workloads when compared to AMD's Zen 4 chips introduced two years ago. Forrest Norrod, executive vice president and general manager of AMD's data center solutions business Group, said the company will "once again take a huge generational leap in performance." CEO Lisa Su emphasized that "AI will enable many new experiences that will make computing an even more essential part of our lives, from accelerating medical discoveries and revolutionizing research to creating smarter, more efficient cities and enabling more resilient supply chains," and stated that AMD's goal is to "make AMD the end-to-end AI leader" by "driving the bleeding edge of performance in CPUs, GPUs and high-performance networking."
Su highlighted dramatic cost advantages for enterprises modernizing their data centers, claiming that one new server running AMD's chip can replace seven legacy servers running Intel chips from four or five years ago, which "significantly reduces the power you need in your data center" and "lowers TCO by more than 60%," allowing enterprises to "break even on their investments in as little as six to 12 months." AMD executives stated they have captured 34% market share in the server CPU market, and analysts affirm the company is executing well against rivals Intel and Nvidia. Jim McGregor, founder and principal analyst at Tirias Research, observed that "AMD just continues to execute year after year. They've gotten to the point where it's just improvement, improvement, improvement," while Ian Cutress, chief analyst of More than Moore, noted "They're hitting all the right corporate notes. They're on track with everything they've talked about," adding that the event demonstrated "their yearly cadence with the GPU, and they're talking about networking and the synergy going in between."
The company deployed two different Zen 5 core architectures for its fifth-generation CPUs: Zen 5, built using the 4-nanometer manufacturing process with up to 128 cores optimized for performance, and Zen 5c, built using 3nm with up to 192 cores designed for efficiency and optimized for parallelization and throughput. This strategy mirrors Intel's approach with performance cores (P-cores) and efficient cores (E-cores), reflecting how data center operators face changing needs with different workload types requiring unique requirements. McGregor explained that "Both Intel and AMD have developed that performance and efficiency core strategy. They realize they have to be more flexible because we've seen some hyperscalers develop their own processors for different applications."
Despite AMD's strong execution, Peter Rutten, research vice president in IDC's worldwide infrastructure research organization, noted that AMD's messaging about delivering a full infrastructure solution including CPUs, GPUs, DPUs, and networking requires the company to beef up its software offerings, as AMD continues to invest in and improve its AMD ROCm software stack.