Intel doesn’t have, nor does it plan to, introduce any kind of chip that could match the performance of what its fiercest competitors, such as NVIDIA, offer for training AI models, according to comments that Intel CEO Pat Gelsinger had reportedly made during a financial conference in August. News of Gelsinger’s statement comes as Intel continues to market Gaudi 3, a cost-efficient AI accelerator that the company launched in September with what is said to be 20% more throughput and 2x price/performance vs. NVIDIA’s H100 GPU for inference of LLaMa 2 70B. Intel also confirmed this week that it would be laying off over 2,000 employees across four states as part of its wider plan in cutting 15,000 jobs: 1,300 employees in Oregon, 385 in Arizona, 319 in California, and 251 in Texas.
At a financial conference in August, Gelsinger admitted that the company isn’t going to be “competing anytime soon for high-end training” because its competitors are “so far ahead,” so it’s betting on AI deployments with enterprises and at the edge. Intel said its strategy for Gaudi 3 accelerator chips will not focus on chasing the market for training massive AI models that has created seemingly unceasing demand for Nvidia’s GPUs…