3 ms·
It's also quite a niche use case... An application fine with low single thread performance, that is highly parallelizable, requiring hundreds of threads, but
by londons_explore 5y ago
It's also quite a niche use case... An application fine with low single thread performance, that is highly parallelizable, requiring hundreds of threads, but with sufficiently branchy execution that CUDA/GPU doesn't work out... Oh, and it also no binary blobs, or you won't be able to port to ARM.
- guiand 5y agoSounds like exactly the use case of a server. Rack space is expensive, condensing it can save the operators money.
- fulafel 5y agoTurning the "vs CUDA" question around, why would you prefer to program a GPU? The dev experience and available progranming languages are worse there plus you are coding to a proprietary hw and sw stack that can't produce portable apps.