According to the intelligence APP, on October 10th local time in the USA, AMD (AMD.US) released the new AI accelerator Instinct MI325X at the "Advancing AI" conference, aiming to compete with Nvidia (NVDA.US) H200. The former is expected to be shipped in the fourth quarter of 2024. At this conference, AMD also introduced the MI350 series chips, which compete with Nvidia's Blackwell series, but the series is not expected to be mass-produced until 2H25.

According to Lisa Su, CEO of AMD, the potential market size (TAM) of data center AI accelerators will grow at a compound rate of over 60% and reach $500 billion by 2028, a significant increase from previous forecasts (Lisa Su previously estimated in early 2024 that it would reach $400 billion by 2027). Overall, this conference did not bring many surprises to investors.

At this meeting, three points were noted: 1) Meta's Llama 405B model is fully running on the MI300X, indicating good progress for AMD in Meta (META.US); 2) AMD did not mention Amazon (AMZN.US) in the customer showcase section; 3) AMD did not disclose the GPU sales targets for 2024 and 2025, as well as the current market supply and demand status.

Highlights of AMD's new GPU: The MI325X accelerator uses a 5-nanometer process and is equipped with 256GB HBM3e memory, with a memory bandwidth of 6TB/s. Its capacity and bandwidth are 1.8 and 1.3 times that of Nvidia's H200, respectively. According to AMD, the theoretical peak performance of MI325X's FP16 and FP8 calculations will both be 1.3 times that of H200. The plan for MI325X to achieve large-scale shipments in 4Q24 is progressing steadily, with the product expected to be launched with server solutions from partners such as Dell (DELL.US), Eviden, Gigabyte, Lenovo, and Super Micro Computer starting in 1Q25. Additionally, AMD also previewed a new generation GPU MI350 accelerator based on the CDNA 4 architecture and 3-nanometer process.

AMD stated that the computational performance (FP16 and FP8) of the MI350X accelerator will increase by 80% compared to MI325X. This product is equipped with 288GB HBM3e memory and supports 8TB/s memory bandwidth. The MI350 series is expected to be launched in 2H25.

Highlights of AMD's new GPU: In addition to releasing new GPU products, AMD also unveiled the fifth generation EPYC "Turin" CPU based on the Zen 5 architecture at this conference. The CPU's performance has been significantly improved compared to the previous generation, especially in data center applications. The Turin CPU has increased the instructions per cycle (IPC) for enterprise and cloud computing workloads by 17%, while the IPC for executing high-performance computing and AI tasks has increased by 37%.

According to AMD, since the launch of the EPYC product line in 2018, AMD has expanded its market share in the global server field from 2% to 34%. AMD has also positioned the EPYC platform as a CPU suitable for both AMD Instinct and Nvidia MGX/HGX platforms. These configurations can support up to 8 OAM MI300X or MI325X GPUs, providing outstanding performance advantages, including a 20% increase in AI inference performance and a 15% increase in training workload capacity. This positions AMD as a key player in the AI CPU field, competing with Intel's Xeon series chips.

AMD's MI325X is a mid-term upgrade of the MI300X, aiming to compete with Nvidia's H200. However, as AMD's next-generation MI350 is planned to be launched in 2H25, AMD will still lag behind Nvidia, considering the latter's B200 will start mass shipping in 4Q24. CICC believes that Nvidia will continue to maintain its leading position in the GPU market, while AMD will continue to strive catching up. In terms of CPUs, AMD's 5th generation EPYC has made significant breakthroughs, gaining more market share in the server field, with its performance and cost-effectiveness superior to Intel's Xeon 6 series.