(Bloomberg) — Intel Corp. Chief Executive Pat Gelsinger took the stage at the Computex trade show in Taiwan to talk about new products that he hopes will help reverse the tide of stock losses for its peers, including the leader AI Nvidia Corp.
Bloomberg’s Most Read
Intel has introduced its new Xeon 6 data center processors with more efficient cores that will allow operators to reduce the space needed for a given task to a third of previous generation hardware. Like rivals from Advanced Micro Devices Inc. to Qualcomm Inc., Intel touted benchmarks that showed its new silicon is significantly better than existing options. The CEOs of AMD and Qualcomm, in previous Computex keynotes, have used Intel’s laptop and desktop processors to show how advanced they are in certain aspects of technology.
Listen to The Big Take daily podcast today.
Gelsinger directly attacked Nvidia CEO Jensen Huang’s claim that traditional processors like Intel’s are losing steam in the age of artificial intelligence.
“Contrary to what Jensen wants you to believe, Moore’s Law is alive and well,” he said, emphasizing that Intel will have an important role to play in the proliferation of AI as a leading supplier of PC chips.
“I think of it like the Internet 25 years ago, it’s that big,” Gelsinger said. “We see this as the fuel that is driving the semiconductor industry to reach $1 trillion by the end of the decade.”
Intel shares were little changed at 9:59 a.m. in New York on Tuesday.
Intel’s Gaudi systems, which compile its chips into multiprocessor kits tailored to handle generative AI training, will be offered by partners including Dell Technologies Inc. and Inventec Corp., Gelsinger said. A kit with eight Intel Gaudi 2 accelerators will sell for US$65,000. A more powerful kit of eight Intel Gaudi 3 accelerators will cost $125,000, with the company estimating both offerings to be more affordable than competitors.
Read more: Nvidia and AMD face off in the fight to take control of AI
Each of these Gaudi 3 clusters is comprised of 8,192 accelerators and Intel estimates that it offers up to 40% faster time to train an AI model compared to an equivalent-sized cluster of Nvidia H100 GPUs. Intel also said that Gaudi 3 would be up to twice as fast as Nvidia’s H100 at performing AI inference tasks, as measured on popular models like those made by Meta Platforms Inc. These advantages may not be enough to take away the Nvidia leads the way in AI processing in data centers.
“The performance of each individual accelerator is no longer the most important thing,” said Leonard Lee, analyst at neXt Curve. Nvidia’s biggest advantage is having a cohesive, integrated ecosystem and proprietary technology like NVLink that ensures its computing clusters work as one. “The power is in being able to create a huge logic accelerator of tremendous size.”
Santa Clara-based Intel has led the computer industry for decades, but its revenue has fallen in the past two years as it has fallen behind rivals. Gelsinger, who was brought back to the company three years ago to turn around its fortunes, has spent heavily to revitalize its offerings and build a network of factories that he says will regain leadership in chip design and manufacturing.
Read more: Intel is spending $28 billion to make Ohio a global chip capital
Although Intel’s sales have stopped declining, analysts are not projecting a quick recovery and the company is on track to end 2024 with $20 billion less revenue than in 2021. Meanwhile, Nvidia’s sales are expected to double and AMD will grow by more than 10 billion. %, according to estimates, as these companies make better use of the flood of spending on AI computing hardware.
“This is the most important moment in our careers together,” said Gelsinger, reiterating the importance of Intel working with its partners. “We were made for this moment.”
–With help from Ian King.
(Updates share price in sixth paragraph.)
Bloomberg Businessweek Most Read
©2024 Bloomberg LP
