Intel: “Microsoft wants to partially run Copilot on the laptop”

According to Intel, Microsoft will soon run parts of Copilot locally on your PC. Only: The minimum requirements for this are higher than what Intel can currently offer.

Microsoft wants to soon run its AI Copilot locally. Intel confirmed this to Tom’s Hardware at the AI ​​Summit. Microsoft Copilot is a generative AI assistant based on the same GPT model that powers OpenAI’s ChatGPT. The copilot can help you with all sorts of daily tasks, like searching emails for action items or preparing PowerPoints. However, all requests to Copilot currently have to go through the cloud, as only there is enough computing power for inference.

Local conclusion

This requires time, network capacity and computing power in the data center. Additionally, privacy-sensitive data may need to be transferred back and forth, and customers don’t always want this. The solution is to run Copilot locally, at least for slightly simpler tasks. This is possible as long as the local hardware has sufficient computing power to process such inference tasks efficiently.

Intel even communicates a number: Microsoft wants 40 tops in NPU computing power. The same number was previously published by Trendforce and we also saw it on slides from HP. Noticeable: Microsoft itself prefers to remain vague about this at the moment and neither wants to confirm nor deny the 40 tops to the editorial team. Fortunately, Intel does this instead of Microsoft.

Future-oriented minimum

40 tops (Tera operations per second) is a measure of the AI ​​performance of an NPU. Currently, no computer that only has one processor on board has enough tops. The NPUs on board Intel and AMD processors have a maximum of ten and sixteen tops. In other words, Copilot cannot run locally on the current generation of “AI PCs,” according to Microsoft’s data, which is now widely confirmed by almost everyone except Microsoft.

Intel is confident that it will reach the lower limit of Copilot performance with the next generation of chips this year. AMD is also undoubtedly aware of the requirements and we would be surprised if new Ryzen chips did not meet the minimum limit.

Microsoft’s plans show where AI is heading. The copilot needs to become ubiquitous using local hardware. In fact, Intel is partly pushing a selling point for Nvidia and AMD: these days, only computers with a separate GPU on board have the necessary tops.