AMD launches a brand-new AI chip that directly targets Nvidia’s GPUs, or data centre graphics processors.
TakeAway Points:
- AMD unveiled a brand-new AI chip that directly targets Nvidia’s GPUs, or data centre graphics processors.
- When the Instinct MI325X is released, it will compete with Nvidia’s Blackwell processors, which will begin to arrive in large quantities early in 2019.
- Nvidia may face price pressure if developers and cloud companies view AMD’s AI chips as a close replacement for their offerings.
AMD’s latest AI chip
AMD’s MI325X chip, will start production before the end of 2024, AMD said Thursday during an event announcing the new product. If AMD’s AI chips are seen by developers and cloud giants as a close substitute for Nvidia’s products, it could put pricing pressure on Nvidia, which has enjoyed roughly 75% gross margins while its GPUs have been in high demand over the past year.
Advanced generative AI such as OpenAI’s ChatGPT requires massive data centers full of GPUs in order to do the necessary processing, which has created demand for more companies to provide AI chips.
In the past few years, Nvidia has dominated the majority of the data center GPU market, but AMD is historically in second place. Now, AMD is aiming to take share from its Silicon Valley rival or at least to capture a big chunk of the market, which it says will be worth $500 billion by 2028.
“AI demand has actually continued to take off and actually exceed expectations. It’s clear that the rate of investment is continuing to grow everywhere,” AMD CEO Lisa Su said at the event.
AMD didn’t reveal new major cloud or internet customers for its Instinct GPUs at the event, but the company has previously disclosed that both Meta and Microsoft buy its AI GPUs and that OpenAI uses them for some applications.
With the launch of the MI325X, AMD is accelerating its product schedule to release new chips on an annual schedule to better compete with Nvidia and take advantage of the boom in AI chips. The new AI chip is the successor to the MI300X, which started shipping late last year. AMD’s 2025 chip will be called MI350, and its 2026 chip will be called MI400, the company said.
The MI325X’s rollout will pit it against Nvidia’s upcoming Blackwell chips, which Nvidia has said will start shipping in significant quantities early next year.
AMD’s stock fell 4% on Thursday. Nvidia shares were up about 1%.
Market challenges
AMD’s biggest obstacle in taking market share is that its rival’s chips use their own programming language, CUDA, which has become standard among AI developers. That essentially locks developers into Nvidia’s ecosystem.
In response, AMD this week said that it has been improving its competing software, called ROCm, so that AI developers can easily switch more of their AI models over to AMD’s chips, which it calls accelerators.
AMD has framed its AI accelerators as more competitive for use cases where AI models are creating content or making predictions rather than when an AI model is processing terabytes of data to improve. That’s partially due to the advanced memory AMD is using on its chip, it said, which allows it to serve Meta’s Llama AI model faster than some Nvidia chips.
“What you see is that the MI325 platform delivers up to 40% more inference performance than the H200 on Llama 3.1,” said Su, referring to Meta’s large language AI model.
