#webgpu

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass ...

Nearly every AI feature in a web app today is a round trip: the page collects text, sends it to a...

How a from-scratch WebGPU LLM engine was checked token by token against llama.cpp, with synthetic models and negative controls.

In 2024, a typical background-removal API round trip looked like this: upload a 2 MB photo, wait...

How SysVis.AI turns natural language and whiteboard sketches into interactive system architecture diagrams running completely offline in your browser via WebGPU.

MicroLLM puts users’ AI keys between their apps and providers, shifting token costs away from developers. Here’s how it works, its security trade-offs, and when browser-based models make more sense.

I've been chasing GPU cycles since machines had 64K of RAM, and every few years the browser quietly...

The Flash Attention 2 path was easier to finish after the algorithm existed in plain JavaScript...

A website may already contain the answers visitors need. The real problem is often finding them. So...

The WebGPU path did not have one honest speedup number. Against multithreaded WASM SIMD, the...

The WebGPU matmul sweep started with a naive kernel, then added 16 by 16 workgroup tiling and a 4 by...

Kernels su misura, fusion delle operazioni e meno accessi in memoria: la ricetta per accelerazioni...