Microsoft Copilot AI Set to Boost Performance Locally on PCs: Intel Reveals Key Details"

Microsoft's Copilot AI service is gearing up to enhance performance by running locally on PCs, according to insights shared by Intel with Tom's Hardware. This advancement signifies a significant shift towards next-gen AI PCs equipped with built-in neural processing units (NPUs) boasting over 40 TOPS (trillion operations per second) of power—a capability currently unmatched by any consumer processor available.

Intel highlights that these AI PCs will be empowered to execute "more elements of Copilot" locally, a departure from the current reliance on cloud processing for even minor tasks. This move aims to reduce latency, potentially enhancing both performance and privacy for users.

Previously rumored requirements for next-gen AI PCs included 40 TOPS of processing power and a modest 16GB of RAM. Presently, Windows lacks extensive NPU utilization, with limited use cases such as background blurring for Surface Studio webcams. In contrast, ChromeOS and macOS leverage NPUs for a broader range of tasks including video and audio processing, OCR, translation, and live transcription.

Despite advancements, the fastest NPU speed is found in Apple's M3 lineup, offering 18 TOPS. Following closely are AMD's Ryzen 8040 and 7040 laptop chips with 16 and 10 TOPS respectively, while Intel's Meteor Lake laptop chips also reach 10 TOPS. Qualcomm's Snapdragon X Elite stands out as a potential candidate for powering Copilot with its 45 TOPS of AI compute speed.

Looking ahead, Intel's Lunar Lake chips slated for a 2025 release promise triple the current NPU speeds. The company recently introduced 300 new AI features optimized for its OpenVino platform and unveiled an AI PC development kit based on the ASUS NUC Pro utilizing current Meteor Lake silicon.

Intel expressed its commitment to advancing AI capabilities across various segments, including desktops, with plans aligning with the evolving landscape of AI PC requirements.