[micro perf] add broadcast op

bytedance / ByteMLPerf

AI Accelerator Benchmark focuses on evaluating AI Accelerators from a practical production perspective, including the ease of use and versatility of software and hardware.

https://bytemlperf.ai/

Apache License 2.0

188 stars 50 forks source link

[micro perf] add broadcast op #62

Closed suisiyuan closed 5 months ago

suisiyuan commented 5 months ago

refer to allreduce op, add broadcast op.

YJessicaGao commented 5 months ago

计算broadcast算子的bus bandwidth时，Correction Factor = 1