Intel has struggled for greater than twenty years to develop a profitable GPU/accelerated computing technique, going all the best way again to the aughts and the ill-fated Larrabee effort. Its most up-to-date efforts centered round Ponte Vecchio and Gaudi chips, neither of which have gained any traction.
Nonetheless, CEO Lip-Bu Tan has made AI a high precedence, reorganizing the company construction round a brand new unit, the AI Group, however that was dealt a setback final November when Sachin Katti, the top of the group, left Intel for a job at OpenAI.
Nonetheless, this reveals Intel is firmly behind GPU/accelerator, says Jon Peddie, president of Jon Peddie Research. “The corporate has been inconsistent in its messaging and its layoff explanations, inflicting pointless hypothesis and confusion,” he mentioned. “His largest problem would be the Intel administration and advertising and marketing machine. However that forms shouldn’t be unknown to him.”
The outcomes of his involvement will most likely not be notable for a yr, says Peddie. “He can’t do a lot in regards to the long-term tasks already underway, however he may also help them keep away from a number of land mines, possibly,” he mentioned.
