# jcjohnson/cnn-benchmarks

Benchmarks for popular CNN models

Repository: https://github.com/jcjohnson/cnn-benchmarks
Canonical: https://ross.abutalabs.com/products/cnn-benchmarks
Language: Python
License: MIT
License Family: permissive
Last push: 2017-09-25T16:58:15+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3703, "days_push": 3264, "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 2532, forks 403 (observed 2026-08-28T04:06:59.014734+00:00)

## What it is
A collection of benchmark results and scripts measuring inference speed of popular CNN models (AlexNet, VGG, ResNet, Inception) across CPUs and various NVIDIA GPUs, with and without cuDNN, using Torch. It includes accuracy and latency tables to help compare hardware and model choices for deep learning.

## Use cases
- compare GPU speed for CNN inference
- decide between GTX 1080 and Titan X for deep learning
- benchmark ResNet vs VGG latency
- measure cuDNN speedup over nn
- find fastest CNN model for my GPU
- evaluate CPU vs GPU training performance

## When to choose
- comparing modern GPUs or frameworks
- you need actively maintained benchmarks
- you work with PyTorch or TensorFlow rather than Torch

## Facets
- artifact type: dataset
- maturity: abandoned
- function: benchmarking, machine-learning, deep-learning
- domain: deep-learning, gpu-computing, performance, computer-vision
- platform: python
- tags: cnn, torch, cudnn, gpu-benchmarks, resnet, vgg, alexnet, inference-speed, linux, gpu

## Member repositories
- jcjohnson/cnn-benchmarks (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:59.014734+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 2026-08-30T02:25:18.361178+00:00, confidence not recorded.
  - readme: https://github.com/jcjohnson/cnn-benchmarks (fetched 2026-08-28T04:06:59.014734+00:00, sha 0991db09a5a2)
- Data as of 2026-08-30T08:39:29.467469+00:00.