In challenge to Nvidia, AMD will ramp up AI chip production this year

At an event in San Francisco, AMD Chief Executive Lisa Su said the new MI300X chip will have 192 gigabytes of memory, more than any current Nvidia chip and a key performance measure for how well the chip will handle the large AI systems behind services similar to ChatGPT. Su said customers will get sample chips in the third quarter, with production rising toward the end of the year.

“There’s no question that AI will be the key driver of silicon consumption for the foreseeable future,,” Su said.

Su also said that a system that combines eight of its MI300X chips into a single computer, competing against similar offerings from Nvidia.

The company also said it has started shipping high volumes of a general purpose central processor chip called “Bergamo” to companies such as Meta Platforms . Alexis Black Bjorlin, who oversees computing infrastructure at Facebook parent Meta, said the firm has adopted the Bergamo chip, which targets a different part of AMD’s data center business that caters to cloud computing providers and other large chip buyers.

Nvidia dominates the AI computing market with 80% to 95% of market share, according to analysts. Last month, Nvidia’s market capitalization briefly touched $1 trillion after the company said it expected a jump in revenue after it secured new chip supplies to meet surging demand.

Discover the stories of your interest

Nvidia shares have surged 170% so far this year and were up 3.3% on Tuesday. AMD stock declined 4% after the announcements but touched a 16-month high on Tuesday and has doubled since the start of the year. Nvidia has few competitors working at a large scale. While Intel Corp and several startups such as Cerebras Systems and SambaNova Systems have competing products, Nvidia’s biggest sales threat so far is the internal chip efforts at Alphabet Inc’s Google and Amazon.com’s cloud unit, both of which rent their custom chips to outside developers.

Nvidia’s lead has come not just from its chips, but from more than a decade of providing software tools to AI researchers and learning to anticipate what they will need in chips that take years to design. AMD on Tuesday provided updates to its Rocm software, which competes against Nvidia’s Cuda software platform.

Soumith Chintala, a Meta vice president who helped create open-source software for artificial intelligence, said during the presentation that he has worked closely with AMD to make it easier for AI developers to use free tools to switch from the “single dominating vendor” of AI chips to other offerings like those from AMD.

“You don’t actually have to do that much work – or almost no work in a lot of cases – to go from one platform to the other,” Chintala said.