AMD announced (according to The Verge) the release of its MI300 chips and Ryzen 8040 mobile processors, set to be available in laptops by 2024, marking a significant step in competing with Nvidia in the AI chip market. The new MI300X accelerators and M1300A APUs are tailored for training and running large language models (LLMs), with 1.5 times more memory capacity than the previous version and improved energy efficiency, aiming to meet the demands of memory and computation for complex LLMs.
AMD CEO Lisa Su highlighted the power of the MI300X, which will be integrated into Azure Virtual Machines in collaboration with Microsoft. Meta also plans to deploy MI300 processors in its data centers. The MI300A APU, set for use in data centers and the El Capitan supercomputer, offers enhanced computing performance and energy efficiency. AMD’s Ryzen 8040 series aims to boost AI processing performance in mobile devices, incorporating neural processing units (NPUs) for more versatile applications like video editing and gaming.
With these developments, AMD is bolstering its position in the competitive AI chip market share race, which Nvidia has dominated with its highly sought-after H100 GPUs. The announcement reflects AMD’s commitment to serving not just cloud providers but also enterprises and startups with powerful AI chips.