umap-gpuGPU-accelerated UMAP

Embed millions of high-dimensional vectors into 2D in seconds — not minutes.

WebGPU Acceleration

The SGD optimization loop runs across thousands of GPU shader cores in parallel. Expect dramatic speedups on large datasets compared to CPU-only UMAP.

HNSW k-NN

Uses hnswlib-wasm for O(n log n) approximate nearest neighbor search — fast and accurate regardless of dataset size.

Transparent CPU Fallback

When WebGPU is unavailable, umap-gpu silently falls back to an identical CPU implementation. Same API, same output, everywhere.