I'm trying to build a VM where I can learn and develop very basic machine learning apps. I'm reading that the ONLY Nvidia card that will work with the ML110 Gen9 is the NVIDIA Quadro M2000. A few questions:
1. Is the NVIDIA Quadro P2000 the next iteration of the M2000?
2. WIll the P2000 work in the Gen9?
3. Is the ML110 so restricted that only ONE GPU will work? I'm almost better off building my own server if that's the case. It would seem that as new GPUs are produced, they would be compatible. Sorta like CPUs. I can slap a newer CPU into the ML110 Gen 9 and it will work. Why don't GPUs work the same? Is it the M2000 or nothing with this particular server?
1. Yes 2. Maybe 3. Actually, you can't install a current Intel CPU into that server, it won't fit/work.
This honestly comes down to the fact that when new GPUs come out, OEM's don't always qualify them on previous servers, they might only qualify them on new servers. Does that mean it won't work? No. It just means HPE didn't invest the time or money into adding more GPU options into that server because the ML350 is the tower server designed for powerful GPUs instead.
I'd say try it out, and if it doesn't work, use the P2000 in a new server you choose. Or if you have budget for a new server, pick up an ML350 and more powerful GPU(s)
