RTX IO in Linux

[johnnynunez]

Hi, We know that you have planned launch rtx IO in windows 11 with the first game in June. But I have another discussion or ideas that I have. A lot of GPU's in University research (PhD) are RTX 3090 because quadro or tesla is a lot of expensive. You know that this graphic card doesn't have professional drivers as the old RTX Titan. My question is... Why is RTX IO not enabled for deep learning?
That what I can only be used now is cmap and is also complicated and we know that directstorage is used by nvidia since 2019 now called magnus IO.
All Turing and Ampere are compatible with RTX IO for gaming. Right now, we all know that the bottleneck is data transfer(batches of data), why couldn't you activate it for young researchers in universities for deep learning?

My request is here:
#17

Official sites:
https://www.nvidia.com/es-es/data-center/magnum-io/
https://www.nvidia.com/en-us/geforce/news/rtx-io-gpu-accelerated-storage-technology/
https://devblogs.microsoft.com/directx/directstorage-api-available-on-pc/
https://developer.nvidia.com/blog/gpudirect-storage/ (2019 news)