Data transfer rate GeForce Quadro

L.s.

I’m wondering if there is any difference between the data transfer rate between “host” and “device” between a GeForce and a Quadron card.
I couldn’t find any hard data regarding this.
At the moment I’m limited by the PCI express-bus and I’m wondering if switchen towards a Quadron card instead of a GeForce card could fix this. So actually if the data transfer rate is higher if you send an image to the card and getting the data back from te card.

Define “limited by PCI-E bus”. What is the bandwidth you’re seeing? What bandwidth do you require?

What are you actually transferring and how?
Maybe you’re not hitting the fastest possible path, e.g. like using OpenGL RGB8 user data to upload to textures instead of BGRA8.

Which API are you using?
If CUDA, have you tried building and running the CUDA bandwidth example to see if you’re maxing out the motherboard capabilities already?
https://docs.nvidia.com/cuda/cuda-samples/index.html#bandwidth-test

See table PCI Express link performance about the theoretical limits here: https://en.wikipedia.org/wiki/PCI_Express
In practice I have not seen any motherboard which reached these limits.
If you’re limited by the PCI-E bus on the motherboard side, it wouldn’t matter what graphics board you used.