Inference › Hardware › GPUs v
4090 vs 5090 for 70B quants

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Linear Mode
4090 vs 5090 for 70B quants
vrambandit Offline
Member
***
Posts: 168
Threads: 14
Joined: Sep 2026
Reputation: 0
#9
09-15-2026, 06:00 AM
Thanks, that helps. The pattern around batch size has been pretty similar for me. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.
2x3090 • llama.cpp • too much VRAM is still not enough
Find
Reply


Messages In This Thread
4090 vs 5090 for 70B quants - by thetarunner - 09-13-2026, 12:06 AM
Re: 4090 vs 5090 for 70B quants - by torchshed - 09-13-2026, 02:48 AM
Re: 4090 vs 5090 for 70B quants - by thetarunner - 09-14-2026, 03:22 AM
Re: 4090 vs 5090 for 70B quants - by batchsize - 09-14-2026, 09:27 PM
Re: 4090 vs 5090 for 70B quants - by amberhall - 09-14-2026, 05:31 PM
Re: 4090 vs 5090 for 70B quants - by cudapoint - 09-14-2026, 05:58 PM
Re: 4090 vs 5090 for 70B quants - by batchwire - 09-17-2026, 02:46 AM
Re: 4090 vs 5090 for 70B quants - by sanevault - 09-14-2026, 05:02 AM
Re: 4090 vs 5090 for 70B quants - by paperport64 - 09-14-2026, 08:13 PM
Re: 4090 vs 5090 for 70B quants - by vrambandit - 09-15-2026, 06:00 AM
Re: 4090 vs 5090 for 70B quants - by thetarunner - 09-15-2026, 02:28 PM
Re: 4090 vs 5090 for 70B quants - by rocmriley - 09-17-2026, 08:05 AM

Forum Jump:


Users browsing this thread: 1 Guest(s)