Intel Gao Yu: XPU solution provides strong local intelligent computing power to drive AI PC landing

It has become a trend that the computing power of the cloud sinks to the side and end. On December 7, the first AI PC Industrial Innovation Forum was held at Lenovo headquarters in Beijing. Gao Yu, general manager of Intel Technology in China, delivered a speech at the forum. He said that Intel AI-ready platform, with CPU+NPU+GPU heterogeneous architecture solution, can provide strong local intelligent computing power for terminals.

Gao Yu, general manager of ▲ Intel China Technology, stressed that Intel is working closely with Lenovo to promote the development and innovation of the AI experience based on the upcoming Meteor Lake processor, helping the perfect deployment of AI functions on PC and speeding up the AI PC landing process.

Gao Yu introduced, based on years of learning and market experience, Intel put forward the concept of XPU, XPU includes GPU, NPU, CPU and so on.

Among them, GPU is characterized by stronger performance and parallel processing ability, which is especially suitable for computing under complex AI load. NPU is a dedicated AI processing unit, which is mainly characterized by low power consumption. At the same time, NPU can also share the load of GPU to achieve better comprehensive performance. CPU is an indispensable important computing power in the platform, which is very suitable for lightweight, low-latency AI tasks. Simple AI applications are very easy to deploy to CPU, which provides great convenience for software development.