Intel unveils game-changing AI chip to challenge Nvidia and AMD

Intel Corp on Monday provided a handful of new details about a chip for artificial intelligence (AI) computing it plans to introduce in 2025 as it shifts strategy to compete against Nvidia Corp and Advanced Micro Devices Inc.

Intel said Monday at a supercomputer conference in Germany that its upcoming “Falcon Shores” chip will have 288 gigabytes of memory and support 8-bit floating-point arithmetic.

These specifications are important because AI models like services like ChatGPT have exploded, and companies are looking for more powerful chips to power them.

The details are also among the first to emerge as Intel implements a shift in strategy to catch up with Nvidia, which leads the AI chip market, and AMD, which is expected to challenge Nvidia’s position with a chip called the MI300.

Intel, on the other hand, has virtually no market share after its potential Nvidia rival, a chip called Ponte Vecchio, suffered years of delays.

Intel announced on Monday that it has almost completed deliveries of Argonne National Lab’s Ponte Vecchio-based Aurora supercomputer, which Intel claims is superior to Nvidia’s latest AI chip, the H100.

But Intel’s Falcon Shores tracking chip won’t hit the market until 2025, when Nvidia will likely get 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 as it abandons its previous strategy of combining graphics processing units (GPUs) with its central processing units (CPUs).

“While we strived to have the best CPU and the best GPU on the market, it was difficult to say that any one vendor at a time had the best combination,” McVeigh told Reuters. “If you have separate offerings that allow you, at the platform level, to choose between both the relationship and the suppliers.”

Read all the Latest Tech News here.