5 ms·
WASM won’t be performant enough for heavy matrix computations because it wouldn’t be able to use SIMD (in its current state). There’s a WASM SIMD proposal in th
by lasagnaphil 5y ago
WASM won’t be performant enough for heavy matrix computations because it wouldn’t be able to use SIMD (in its current state). There’s a WASM SIMD proposal in the works but it’s only for 128-bit vectors and I’m not sure it will perform the same as native code.