The Evolution of x86 SIMD: From SSE to AVX-512
bgslabs.orgThis was already posted: https://news.ycombinator.com/item?id=47047936
It contains many factual errors.
Hey author here. I have actually rewritten the parts where you might call AI slop. I have tried to correct the text to the best of my abilities.
https://bgslabs.org/blog/evolution-of-x86-simd/#acknowledgem...
If you look at there I tried to be as transparent about this process as possible. I simply didn't know any better than to use AI to fact check my data when I first started - which was a really bad idea and led to the horrendous outcome as you've seen there. I am not trying to hide anything, I made a mistake. If you could give the article a re-read and tell me where I might have gone wrong I would be really happy. I actually want this to be a good and useful educational resource, not AI slop.
Thank you for your time regardless.
I don't know much about this, but I'm wondering things:
- how do ARM's NEON/SVE fit in in this history?
- is there a AVX/SVE SIMD equivalent being standardized across GPU vendors?
> standardized across GPU
Nothing is getting standardized across GPU vendors never ever again (pretty much every GPU uses some form of SIMD internally). Nvidia's CUDA is too specific to Nvidia's hardware, AMD have some serious problem delivering software/API part, Intel could but has too low of a market share, the rest are mobile IP core vendors with zero interest to provide compatibility.
Excellent