Parallelization SIMD support like un JuliaLang

[acostaja]

Please add support to vectorial (parallelized) instructions with something like in Julia Lang with it's "." Dot-syntax
This opens opportunities in HPC/GPGPU/ML where this is very useful for SIMD optimization. Also include some syntax/keyword to enable such features being complied like declaration to CUDA << >> non STD C syntax being generated when compiled/translated to/from CUDA, HIP/opencl

[medvednikov]

Yes, this would be nice.

[dumblob]

Any special syntax for SIMD parallelization will fall short rather quickly (and has too many shortcomings). From my experience the best way is to stick with intrinsics (as mentioned in #1868 (comment) ) as this approach is by definition extendable and because they already exist e.g. in C++, there is no need to reinvent the wheel.

See also an older issue #1839 .