shubuZ
Re:Triton: Open-Source GPU Programming for Neural Networks
on Wednesday July 28, 2021 @01:09PM (#27986188)
I have found writing CUDA code is much simpler than writing correct multi-threaded AVX2/AVX-512 code.
Re:Triton: Open-Source GPU Programming for Neural Networks
on Wednesday July 28, 2021 @01:07PM (#27986176)
Which other hardware vendor provides the level of performance that Nvidia's GPU provide?
Wasnt the benevolence on single (or couple) manufacturer(s) true in 90s, 2020s?
Re:Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU
on Friday May 15, 2020 @12:40PM (#23194183)
The network topology is the same, no modules are removed, etc. Of course it will be optimized to run on Ampere