Alibaba XuanTie C910

Realizes: Alibaba cloud/edge inference scenarios

The 2020 vector-plus-scalar Alibaba XuanTie C910 couples a wide vector unit with a scalar pipeline to enable edge and AI acceleration workloads.

Examples

🔗

Alibaba Pingtouge edge server

Edge AI inference on the Alibaba Pingtouge board powered by the vector-plus-scalar C910 core.

AI inference with pulse throughput metrics for vision and language workloads on an Alibaba Pingtouge edge server. latency-critical edge 3.2 nJ per inference pulse