Huawei Cloud releases the FlexNPU flexible intelligent computing operating system.

date
21/03/2026
On March 20th, at the Huawei Cloud small and medium-sized enterprise AI solution launch event, the flexible intelligent computing operation system FlexNPU was introduced. It aims to enhance the effective utilization of computing power in the inference pool by building powerful and flexible AI infra layer computational scheduling technology. This will ultimately provide a greater token throughput output with the same AI hardware computing power investment.