Intel on Monday provided a handful of new details about an artificial intelligence (AI) chip it plans to introduce in 2025 as it shifts its strategy to compete with Nvidia and Advanced Micro Devices.
At a supercomputing conference in Germany on Monday, Intel said its forthcoming “Falcon Shores” chip will have 288 GB of memory and support 8-bit floating-point arithmetic. Those tech specs are important as artificial intelligence models, similar to services like ChatGPT, have exploded in size and companies are looking for more powerful chips to run them.
The details are also among the first to trickle out as Intel makes a strategy change to overtake Nvidia, which leads the market in chips for AI, and AMD, which is expected to challenge Nvidia’s position with a chip called the MI300.
Intel, on the other hand, has essentially no market share after its future Nvidia competitor, a chip called Ponte Vecchio, was delayed for years.
Intel said Monday it has nearly completed deliveries of Argonne National Lab’s Aurora supercomputer, based on the Ponte Vecchio, which Intel says has better performance than Nvidia’s latest AI chip, the H100.
But Intel’s Falcon Shores follow-up chip won’t hit the market until 2025, when Nvidia will likely have another chip of its own.
Jeff McVeigh, interim head of Intel’s accelerated computing systems and graphics group, said the company is taking time to rework the chip after abandoning its previous strategy of combining graphics processing units (GPUs) with its central processing units (CPUs) .
“While we strive to have the best CPU and the best GPU on the market, it was hard to say that one vendor at a time would have the best combination of those,” McVeigh told Reuters. “If you have discrete offerings, you can choose between both the ratio and the suppliers at the platform level.”
© Thomson Reuters 2023
