search

mlcommons / inference_results_v0.5

This repository contains the results and code for the MLPerf™ Inference v0.5 benchmark.
https://mlcommons.org/en/inference-datacenter-05/
Apache License 2.0
55 stars 43 forks source link

Habana: 40 CPUs in open submissions #8

Open psyhtest opened 4 years ago

psyhtest commented 4 years ago

Habana's open submissions claim to have 40 CPUs: low latency, medium latency. Habana's closed submissions seem to refer to the same dual-socket system and correspondingly 2 CPUs.

/cc @itayhubara