# baidu-research/DeepBench

Benchmarking Deep Learning operations on different hardware

Repository: https://github.com/baidu-research/DeepBench
Canonical: https://ross.abutalabs.com/products/deepbench
Language: C++
License: Apache-2.0
License Family: permissive
Last push: 2021-04-25T09:13:47+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3631, "days_push": 1956, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1106, forks 242 (observed 2026-08-28T04:03:36.560648+00:00)

## What it is
DeepBench is a benchmarking suite from Baidu Research that measures the performance of fundamental deep learning operations (matrix multiplication, convolutions, etc.) across different hardware platforms. It targets low-level operations suitable for hardware evaluation rather than full model training.

## Use cases
- compare gpu performance for deep learning operations
- benchmark matrix multiplication on different hardware
- evaluate hardware for neural network training
- measure convolution kernel performance
- test hardware suitability for deep learning inference
- profile low-level deep learning ops for chip design

## When to choose
- you need to compare raw operation-level performance across GPUs or CPUs
- you are evaluating or designing hardware for deep learning workloads
- you want to identify bottlenecks in neural network libraries like cuDNN or MKL

## When to avoid
- you want to benchmark end-to-end training time of full models
- you need application-level or framework-level benchmarks
- you need a tool that is actively updated with the latest hardware support

## Facets
- artifact type: library
- maturity: maintenance
- function: benchmarking, gpu-computing, deep-learning
- domain: deep-learning, gpu-computing, performance, hardware
- platform: cpp, cross-platform
- tags: neural-network-kernels, hardware-evaluation, matrix-multiplication, convolution, training-benchmark, inference-benchmark, linux, gpu

## Member repositories
- baidu-research/DeepBench (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:36.560648+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:44:13.532409+00:00, confidence not recorded.
  - readme: https://github.com/baidu-research/DeepBench (fetched 2026-08-28T04:03:36.560648+00:00, sha 1c7b01b762ed)
- Data as of 2026-08-30T08:39:29.467469+00:00.