Intel Promises 40 TOPS AI Chips to Power Local Copilot on Next-Gen PCs
-
Intel confirms Microsoft's Copilot AI will soon run locally on PCs instead of the cloud to provide latency, performance, and privacy benefits.
-
Next-gen AI PCs will require 40 TOPS of performance from the Neural Processing Unit (NPU). Intel's current chips offer 10-16 TOPS.
-
Running Copilot locally frees up the GPU and CPU for other tasks. Microsoft wants it run on NPU to minimize battery drain.
-
Intel promises next-gen Lunar Lake chips later this year with 3x more AI performance, as part of a product roadmap to address all segments of the AI market.
-
There will likely be a "TOPS war" between Intel, AMD, and Qualcomm to reach the 40 TOPS requirement for next-gen AI PCs.