Pinned Loading

1. zero-tvm zero-tvm Public

   Phi-3-mini decoding in a browser on 10 hand-written WGSL kernels (27 files, 3,078 LOC). 228 dispatches/token at ~40 tok/s on M2 Pro — 22% behind WebLLM's TVM-autotuned 85 kernels on identical weights.

   WGSL 2 2

2. webgpu-kernel-fusion webgpu-kernel-fusion Public

   Research umbrella for GPU kernel fusion — eliminating per-dispatch overhead in sequential GPU workloads, from RL to transformer inference.

   JavaScript 3

3. gpubench gpubench Public

   Real WebGPU compute benchmarks in the browser. 5 compute workloads + fused transformer decoding across 4 kernel variants — 159–720× over PyTorch on the same GPU.

   TypeScript

4. markview markview Public

   The markdown viewer your docs deserve — beautiful rendering, full-text search, 15 MCP tools, presentation mode, PWA. Privacy-first.

   TypeScript

5. webgpu-dna webgpu-dna Public

   WebGPU port of Geant4-DNA: in-browser Monte Carlo track-structure simulation with IRT chemistry and B-DNA SSB/DSB scoring

   WGSL

6. webgpu-q webgpu-q Public

   WebGPU quantum many-body simulator — statevector + MPS + kernel fusion + chemistry. 160 tests, ITensor-validated, PySCF to 7 decimals. Browser-native.

   TypeScript