I’m looking for some GPUs for our lab’s cluster. We need GPUs to do deep learning and simulation rendering. We feel a bit lost in all the available models and we don’t know which one we should go for.
This article says that the best GPUs for deep learning are RTX 3080 and RTX 3090 and it says to avoid any Quadro cards. Is this true?
If anybody could help me with choosing the right GPU model for our cluster, I would greatly appreciate it.
Our system is composed of 28 nodes that run Ubuntu 20.04. We would like to use CUDA with pytorch for deep learning models and we do a lot of robotics simulations.
Thanks in advance for your help.
it is a question of budget, by the way why PyTorch compared to Keras? , there is little difference between a 3080 and a 3090, given the number of nodes, I would go for 3060 Ti, would only be for power consumption