Intel used a supercomputing conference in Germany to reveal that its forthcoming "Falcon Shores" chip will have 288 gigabytes of memory and support 8-bit floating point computation; these would provide the necessary power and performance to support artificial intelligence models similar to services like ChatGPT.
Intel is looking to catch-up to Nvidia, which leads the market in chips for AI, and AMD, which has launched a new AI-focused chip called the MI300. Intel’s Ponte Vecchio, a chip originally intended to compete with Nvidia, has suffered years of delays and left the computing giant with little market share in the fast-growing AI market.
However, Intel's Falcon Shores chip won't be ready to go to market until 2025, by which time Nvidia is expected to have developed and delivered another chip.
Commenting Jeff McVeigh, corporate vice president of Intel's super compute group, said the company had decided to take time to rework the chip after giving up on its prior strategy of combining graphics processing units (GPUs) with its central processing units (CPUs).
"While we aspire to have the best CPU and the best GPU in the market, it was hard to say that one vendor at one time was going to have the best combination of those," McVeigh told Reuters. "If you have discrete offerings, that allows you at the platform level to choose both between the ratio as well as the vendors."
