Shares of AMD rose almost 5% Thursday, a day after Microsoft’s technology chief said the chipmaker is bolstering its position in artificial intelligence, where Nvidia dominates.
So far this year, Nvidia shares have almost tripled while AMD is up about 60%. Since the launch in late 2022 of OpenAI’s ChatGPT chatbot, the tech industry has been swarming to new large language models, which require hefty processing power.
Nvidia’s graphics processing units are handling so much of those workloads that the company is forecasting 170% year-over-year revenue growth in the current quarter. AMD announced in June that during the third quarter, it would start sampling its MI300X chip with clients. Those GPUs were designed specifically for AI models.
“They’re making increasingly compelling GPU offerings that I think are going to become more and more important to the marketplace in the coming years,” Kevin Scott, Microsoft’s chief technology officer, said at the Code Conference in Dana Point, California, on Wednesday.
Microsoft and AMD are longtime partners, and it’s in Microsoft’s interest to have more high-powered chips on the market from a broader set of vendors. For years, Microsoft has offered some AMD GPUs to its Azure cloud customers, in addition to powering some of its computers and its Xbox consoles with AMD chips.
In May, AMD said Microsoft had started offering a cloud networking service to clients, drawing on the chipmaker’s Pensando products.
At Code, The Verge’s Nilay Patel asked Scott how easy it would be to adopt AMD’s GPUs at scale and move away from Nvidia. Scott declined to answer directly, saying that developers using the AI programming tools shouldn’t need to think about the hardware under the hood.
Scott did note that “competition is certainly a very good thing.”
Bloomberg reported in May that AMD was working with Microsoft on a custom AI chip, but Scott declined to say if that’s actually happening. Microsoft’s cloud rivals Amazon and Google have developed homegrown silicon.
