You can do slightly better fairly easily I think. See here for example https://g...

		coffeeaddict1 on Nov 11, 2024 \| parent \| context \| favorite \| on: Optimizing a WebGPU Matmul Kernel for 1 TFLOP You can do slightly better fairly easily I think. See here for example https://github.com/AnswerDotAI/gpu.cpp/pull/35