3 ms·
> [WASM is] still painfully slow compared to AVX or Neon enabled assembly. There's an open proposal to add SIMD intrinsics to WASM. https://github.com/WebAsse
by duskwuff 5y ago
> [WASM is] still painfully slow compared to AVX or Neon enabled assembly.
There's an open proposal to add SIMD intrinsics to WASM.
https://github.com/WebAssembly/proposals/issues/1 https://github.com/WebAssembly/proposals/issues/1
- charcircuit 5y agoThat's an issue from 2018 with no replies to it.
- azakai 5y agoFor current status ("phase 3" as that link says is maybe not obvious enough), there is a solid spec and multiple implementations, see the "Fixed-width SIMD" line here: https://webassembly.org/roadmap/ https://webassembly.org/roadmap/ And here is an overview of this spec: https://github.com/WebAssembly/simd/blob/master/proposals/simd/SIMD.md https://github.com/WebAssembly/simd/blob/master/proposals/si...
- the_duke 5y agoThe fixed width SIMD proposal was standardized a while ago and is available in quite a few runtimes. https://github.com/WebAssembly/proposals/blob/main/finished-proposals.md https://github.com/WebAssembly/proposals/blob/main/finished-...
- ashvardanian 5y agoYeah, saw it, but never tried. Java, C# and other high level languages have libraries for basic SIMD operations. It's completely different world from manually invoking intrinsics in C or writing raw assembly. I am wondering where WASM will land. Visually it's definitely not Assembly. It actually looks more like Python than C. I guess my subjective assumptions are based entirely on looks. Will have to try it for smth, will definitely share afterwards :)