Today Baidu Research announced DeepBench, a benchmarking tool for evaluating the performance of deep learning operations across hardware platforms. The announcement was made at the O’Reilly AI Conference by Greg Diamos and Sharan Narang of Baidu Research’s Silicon Valley AI Lab.
Deep learning developers and researchers want to train neural networks as fast as possible. Right now we are limited by computing performance,” said Dr. Diamos. “The first step in improving performance is to measure it, so we created DeepBench and are opening it up to the deep learning community. We believe that tracking performance on different hardware platforms will help processor designers better optimize their hardware for deep learning applications.”
Initially, DeepBench is providing benchmark results for four platforms: NVIDIA TitanX, NVIDIA M40, NVIDIA TitanX Pascal and Intel Xeon Phi processors. Other hardware vendors are invited to submit results.
“DeepBench will help organizations — from startups to large companies — identify the optimal hardware platform for their deep learning applications,” added Mr. Narang. “Researchers can contribute new operations to DeepBench as they develop new algorithms, which will provide them with a voice to communicate their performance needs to hardware designers.”