# DmitryUlyanov/Multicore-TSNE

Parallel t-SNE implementation with Python and Torch wrappers.

Repository: https://github.com/DmitryUlyanov/Multicore-TSNE
Canonical: https://ross.abutalabs.com/products/multicore-tsne
Language: C++
License: NOASSERTION
License Family: other
Topics: barnes-hut-tsne, multicore, py-bh-tsne, tsne
Last push: 2024-02-06T10:59:55+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": 3605, "days_push": 939, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1909, forks 229 (observed 2026-08-28T04:05:53.044837+00:00)

## What it is
A multicore modification of Barnes-Hut t-SNE written in C++ with Python (CFFI) and Torch wrappers for fast dimensionality reduction of high-dimensional data. It is faster than sklearn's t-SNE even on a single core and scales with available CPU cores.

## Use cases
- visualize high-dimensional datasets like MNIST in 2D with t-SNE
- speed up t-SNE embedding of large datasets using multiple cores
- run t-SNE on a 300k x 3000 feature table quickly
- replace sklearn TSNE with a faster Barnes-Hut implementation
- embed features for a Kaggle competition under a deadline
- compute t-SNE embeddings from Torch tensors

## When to choose
- you need t-SNE on large high-dimensional data and sklearn is too slow
- you have many CPU cores and want parallel nearest-neighbor probability computation
- you want a Python or Torch wrapper around a fast C++ Barnes-Hut t-SNE

## When to avoid
- your data has very low dimensionality, where parallelization gains are minimal
- you need actively maintained software with recent development
- you need UMAP or other scalable dimensionality reduction alternatives instead of t-SNE

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, data-visualization, math
- domain: machine-learning, data-visualization, data-science
- platform: python, cpp, cross-platform
- tags: tsne, barnes-hut, dimensionality-reduction, multicore, parallel-computing, embedding

## Member repositories
- DmitryUlyanov/Multicore-TSNE (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:53.044837+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 2026-08-30T03:11:05.693531+00:00, confidence not recorded.
  - readme: https://github.com/DmitryUlyanov/Multicore-TSNE (fetched 2026-08-28T04:05:53.044837+00:00, sha a22ca4d77127)
- Data as of 2026-08-30T08:39:29.467469+00:00.