It seems that Zig is well suited for writing SIMD [0]. If only GPU makers could standardise an extended ISA like AVX on CPU, and we could all run SIMD or SIMT

It seems that Zig is well suited for writing SIMD [0]. If only GPU makers could standardise an extended ISA like AVX on CPU, and we could all run SIMD or SIMT code without needing any librairies, but our compilers. [0]: https://zig.guide/language-basics/vectors/ https://zig.guide/language-basics/vectors/