4 ms·
I agree about GPUs. Intel used to have a low-end CPU with a 128MB L4 cache that was shared with the iGPU. It turns out having that large cache improved a lot
by CoolGuySteve 6y ago
I agree about GPUs. Intel used to have a low-end CPU with a 128MB L4 cache that was shared with the iGPU. It turns out having that large cache improved a lot of regular software too. I'd rather see die space used for that than AVX512.
The only AVX512 instructions that actually seem essential imo are the VNNI extensions.
There are a lot of use cases for a neural network that is trained on the GPU but where inference is done on a bunch of CPUs because it's cheaper than buying/renting inference hardware.
- cogman10 6y agoI think you are thinking AVX takes up more diespace than it does. Only a tiny fraction of all CPU diespace is dedicated to the actual CPU. The rest goes primarily to cache and partially to the various controllers on the CPU (memory controller, PCI-Express controller, etc). There's a reason AMD is and Intel are able to cram full GPUs onto the same die as the CPU.
- Dylan16807 6y agohttps://imgur.com/r/intel/4SG36 https://imgur.com/r/intel/4SG36 Yeah, here's a good image of how much space it costs to add a 'second' AVX-512 unit to certain SKUs. Cutting AVX-512 support entirely would save a bit more space, and the total would be enough to make a difference, but it's not an extreme amount. https://software.intel.com/sites/default/files/managed/5f/99/Skylake_core.jpg https://software.intel.com/sites/default/files/managed/5f/99... https://twitter.com/GPUsAreMagic/status/1256866465577394181 https://twitter.com/GPUsAreMagic/status/1256866465577394181 these too (speculative)
- jagger27 6y agoAnd the GPUs are usually the bigger component.