# Bringing UMAP Closer to the Speed of Light with GPU Acceleration

arXiv: https://arxiv.org/abs/2008.00325 · HTML: https://mlsanity.com/paper/2008.00325/

## Key facts

- Cited by 19 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 1 Aug 2020.

## Built on by (most cited first, 19 of 19)

- [From t-SNE to UMAP with contrastive learning](https://mlsanity.com/paper/2206.01816/) (2022)
- [CAGRA: Highly Parallel Graph Construction and Approximate Nearest Neighbor Search for GPUs](https://mlsanity.com/paper/2308.15136/) (2023)
- [Uniform Manifold Approximation with Two-phase Optimization](https://mlsanity.com/paper/2205.00420/) (2022)
- [Holographic Global Convolutional Networks for Long-Range Prediction Tasks in Malware Detection](https://mlsanity.com/paper/2403.17978/) (2024)
- [GPU Semiring Primitives for Sparse Neighborhood Methods](https://mlsanity.com/paper/2104.06357/) (2021)
- [cuSLINK: Single-linkage Agglomerative Clustering on the GPU](https://mlsanity.com/paper/2306.16354/) (2023)
- [Federated t-SNE and UMAP for Distributed Data Visualization](https://mlsanity.com/paper/2412.13495/) (2024)
- [UMATO: Bridging Local and Global Structures for Reliable Visual Analytics with Dimensionality Reduction](https://mlsanity.com/paper/2508.16227/) (2025)
- [Scalable semi-supervised dimensionality reduction with GPU-accelerated EmbedSOM](https://mlsanity.com/paper/2201.00701/) (2022)
- [Tracking Discourse Influence in Darknet Forums](https://mlsanity.com/paper/2202.02081/) (2022)
- [GiDR-DUN; Gradient Dimensionality Reduction -- Differences and Unification](https://mlsanity.com/paper/2206.09689/) (2022)
- [Lens functions for exploring UMAP Projections with Domain Knowledge](https://mlsanity.com/paper/2405.09204/) (2024)
- [Out-of-Core Dimensionality Reduction for Large Data via Out-of-Sample Extensions](https://mlsanity.com/paper/2408.04129/) (2024)
- [Open-source framework for detecting bias and overfitting for large pathology images](https://mlsanity.com/paper/2503.01827/) (2025)
- [NOMAD Projection](https://mlsanity.com/paper/2505.15511/) (2025)
- [A Vector Symbolic Approach to Multiple Instance Learning](https://mlsanity.com/paper/2511.16795/) (2025)
- [Manifold Bandits: Bayesian Curriculum Learning over the Latent Geometry of Large Language Models](https://mlsanity.com/paper/2606.19750/) (2026)
- [Gleam: Adaptive Network-Efficient CUDA API Remoting for Cross-Device GPU Sharing over LANs](https://mlsanity.com/paper/2607.23115/) (2026)
- [ibUMAP: Coherent and Scalable Field Evaluation for UMAP Optimization](https://mlsanity.com/paper/2610.01445/) (2026)

## Abstract (arXiv)

The Uniform Manifold Approximation and Projection (UMAP) algorithm has become widely popular for its ease of use, quality of results, and support for exploratory, unsupervised, supervised, and semi-supervised learning. While many algorithms can be ported to a GPU in a simple and direct fashion, such efforts have resulted in inefficient and inaccurate versions of UMAP. We show a number of techniques that can be used to make a faster and more faithful GPU version of UMAP, and obtain speedups of up to 100x in practice. Many of these design choices/lessons are general purpose and may inform the conversion of other graph and manifold learning algorithms to use GPUs. Our implementation has been made publicly available as part of the open source RAPIDS cuML library (https://github.com/rapidsai/cuml).