{"library":"cuda-wasm","type":"library","category":null,"description":"cuda-wasm (v1.1.1) is a high-performance transpiler that converts CUDA C++ code to WebAssembly and WebGPU, enabling GPU-accelerated computing in browsers and Node.js without NVIDIA hardware. It uses a clean-room implementation based on public CUDA specifications, translating CUDA syntax to Rust/WebGPU patterns. Ships with TypeScript types, supports Node >=16.0.0. Key differentiators: no NVIDIA dependencies, near-native performance via WebGPU, ruv-FANN neural network integration. Released under MIT/Apache-2.0. Active development.","language":"javascript","status":"active","version":"1.1.1","tags":["javascript","cuda","webassembly","wasm","webgpu","gpu","parallel-computing","transpiler","rust","typescript"],"last_verified":"Fri May 01","install":[{"cmd":"npm install cuda-wasm","imports":["import { cudaWasm } from 'cuda-wasm'","import { transpileCUDA } from 'cuda-wasm/transpile'","import { runKernel } from 'cuda-wasm/runtime'","import { NeuralNetwork } from 'cuda-wasm/neural'"]},{"cmd":"yarn add cuda-wasm","imports":[]},{"cmd":"pnpm add cuda-wasm","imports":[]}],"homepage":null,"github":"https://github.com/ruvnet/ruv-FANN","docs":null,"changelog":null,"pypi":null,"npm":"https://www.npmjs.com/package/cuda-wasm","openapi_spec":null,"status_page":null,"smithery":null,"compatibility":null}