# mit-han-lab/torchsparse

[MICRO'23, MLSys'22] TorchSparse: Efficient Training and Inference Framework for Sparse Convolution on GPUs.

Repository: https://github.com/mit-han-lab/torchsparse
Canonical: https://ross.abutalabs.com/products/torchsparse
Homepage: https://torchsparse.mit.edu
Language: Cuda
License: MIT
License Family: permissive
Topics: pytorch, acceleration
Last push: 2025-02-24T13:52:53+00:00

## Health v2 (maintenance only)
Score: 26/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 8, release rhythm 8, longevity 100
- inputs: {"age_days": 2200, "days_push": 555, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1472, forks 192 (observed 2026-08-28T04:04:49.469635+00:00)

## What it is
TorchSparse is a high-performance PyTorch library for sparse convolution on 3D point clouds, with optimized GPU kernels for both training and inference. TorchSparse++ achieves significant speedups over prior sparse convolution libraries like MinkowskiEngine and SpConv.

## Use cases
- run sparse convolutions on 3D point clouds fast
- train point cloud segmentation networks on GPU
- accelerate LiDAR perception models for autonomous driving
- speed up 3D object detection inference
- replace MinkowskiEngine or SpConv with a faster backend
- integrate sparse convolution into MMDetection3D or OpenPCDet
- 3D mesh and surface reconstruction from point clouds

## When to choose
- you need state-of-the-art sparse convolution performance on NVIDIA GPUs
- you are building LiDAR or point cloud perception pipelines
- you want to plug into MMDetection3D or OpenPCDet
- you need both training and inference support for sparse CNNs

## When to avoid
- you work with dense 2D images rather than sparse 3D data
- you need CPU-only or non-NVIDIA GPU support
- you want a turnkey application rather than a low-level library
- your project depends on a different sparse conv library's exact API

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, gpu-computing, image-processing
- domain: computer-vision, deep-learning, autonomous-vehicles, gpu-computing
- platform: python
- tags: sparse-convolution, point-cloud, pytorch, 3d-deep-learning, lidar, gpu-acceleration, autonomous-driving, gpu, linux, cuda

## Member repositories
- mit-han-lab/torchsparse (main) score 26

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:49.469635+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 2026-08-30T04:34:42.275364+00:00, confidence not recorded.
  - readme: https://github.com/mit-han-lab/torchsparse (fetched 2026-08-28T04:04:49.469635+00:00, sha b4659654152a)
  - homepage: https://torchsparse.mit.edu (fetched 2026-08-29T11:42:24.564779+00:00, sha 01654bf03d67)
  - site_page: https://hanlab.mit.edu/ (fetched 2026-08-29T11:42:24.574305+00:00, sha 4be659389a86)
- Data as of 2026-08-30T08:39:29.467469+00:00.