Intel Corp. has some major opportunities in the arms race to power artificial intelligence, but its ability to hit the gas will be key.
Executives talked up AI effusively on Intel’s INTC,
See: Intel stock rallies after earnings show AI data-center beat, strong PC sales
But the pressure is on and competition is stiff. Intel has a recent history of blunders, but it will need to execute on product launches and not slip with delays, as happened for a brief period before Gelsinger took over.
Intel also has a lot on its plate, as it’s in the throes of creating a contract manufacturing business that it hopes will compete directly with Taiwan Semiconductor Manufacturing Co. TSM,
While the opportunities in AI are massive, Intel will need to defy some doubters in order to achieve success.
Ruben Roy, a Stifel & Co. analyst, wrote in a recent note that he believed Intel rival Advanced Micro Devices Inc. AMD,
Gelsinger did not specify at what point next year the new Gaudi3 AI chip is expected, but he said Intel will have some “very positive updates there in the future with both customers as well as expanded business opportunities.”
Many companies with data centers, cloud-service providers and hyperscalers are working frantically to add more servers with higher-end chips to handle the onslaught of computing power needed to run AI, such as training and inference, and the new slew of generative-AI applications being developed. But so far Nvidia Corp. NVDA,
Back in May, Nvidia gave a stunning quarterly forecast for what would be a record $11 billion in revenue, based in large part on surging demand for its AI-related chips.
Gelsinger also acknowledged some near-term “wallet focus” on AI accelerator chips, versus the general purpose CPU chips that Intel sells more of, which has been an area of concern on Wall Street. In that sense, Intel expects to see server CPUs modestly decline on a sequential basis in the third quarter before showing recovery in the fourth quarter.
Beyond in the data center, Intel could also capitalize on huge potential for AI in the PC market. Gelsinger said the company’s next Meteor Lake chips will be “ushering in the AI PC generation,” and he likened the process of bringing AI to the PC to the company’s past Centrino chip family that brought WiFi functionality to the PC back in 2003.
In Intel’s view, AI-enabled PC chips will be critical due to the latency challenges brought upon by the process of going back and forth to the cloud or data centers.
“You can’t round-trip to the cloud if you don’t have the latency, the bandwidth or the cost structure,” Gelsinger said. “With that in mind, we do see this idea of bringing AI directly into the client and Meteor Lake, which we’re bringing into the market in the second half of the year.”
He said it will be the first major product for the PC that includes native AI capabilities, and the neural engine Intel has talked about. “This will be a volume delivery that we will have, and we expect that Intel is the volume leader for the client footprint,” he said.
Intel certainly made a case for its place in the AI revolution, but it remains to be seen whether Gelsinger can avoid his predecessors’ mistakes in getting the company to its destination — all while rivals chase the same juicy market.
The clock is ticking.