In AVX512 and Vec512 · Issue #56187 · pytorch/pytorch · GitHub we are considering dropping AVX support. This means to get vectorized CPU kernels, you must have a CPU recent enough to support AVX2, otherwise you will get unvectorized operations. If this would unduly affect you, please let us know on the issue.
Do we have a way to get some statistics about that via PyPi or conda by any chance?
Probably not but there was some steam statistics and it was 82.28% with AVX2 support (and its growing by like ~0.5% month to month) and 94.77% with AVX support. So i would think percentage should be lot higher in ML based builds 
While it’s not an issue for researchers/data scientists/cloud services, it can be an issue when you need to use PyTorch inference within softwares meant for consumers.
For instance macOS 10.14 and 10.15 (still supported by Apple) can run on CPUs without AVX2 instructions. Only macOS 11.0 requires AVX2 CPUs.
That being said, it’s not dramatic if PyTorch then fallback to a non-vectorized path. It would be problematic if it couldn’t run at all without AVX2.
Yes, what would happen in those cases is it would just degrade to nonvectorized.
Hi, I was just wondering if PyTorch will continue to support non-AVX2 CPUs in the coming years, or if you’ve been contemplating on sneaking in a hard requirement. I can live without non-vectorization because the performance of my PyTorch-dependent projects is GPU-limited (CUDA) anyway.
I really do not want to buy a new machine ever since the AI / data center industry caused the prices of DDR4/DDR5 RAM and SSD to at least triple over the past year. Time will tell, but this does not seem like a short blip like the crypto craze from late 2017 to early 2018 which caused for example the Radeon RX 580 to cost $600 at its peak. My i7-3770K and 32GB of DDR3 work perfectly fine for my purposes.