Using Gerforce GTX 1080 Ti PNY and INNO3D together

[williepatriot]

I have a computer with an 11GB GTX 1080Ti PNY and just got a new 11GB GTX 1080Ti INNO3D. Can I have both of them work on the same computer? Or do they need to be exactly the same model?

 

[Space Lynx]

I have a computer with an 11GB GTX 1080Ti PNY and just got a new 11GB GTX 1080Ti INNO3D. Can I have both of them work on the same computer? Or do they need to be exactly the same model?

they don't need to be the same model.


they will work together fine. don't forget to update your drivers!

 

[cucker tarlson]

yup but remember that if you're using the hard HB bridge they have to be the same width. the bridge must go all the way on the sli fingers.

 

[qubit]

They should work together just fine as branding isn't supposed to matter any more. You might also want to apply the NVIDIA DisplayPort fix on both cards.

http://www.nvidia.com/object/nv-uefi-update-x64.html

 

[williepatriot]

yup but remember that if you're using the hard HB bridge they have to be the same width. the bridge must go all the way on the sli fingers.

Do you know if SLI is necessary for Deep Learning (i.e. parallel GPU computing, no Graphics)? Sadly, I probably won't be gaming on this computer, but I read somewhere that SLI wouldn't make any difference for training models in e.g. Tensorflow or Keras.

 

[cucker tarlson]

No you don't have to have sli for compute.

https://blog.inten.to/hardware-for-deep-learning-part-3-gpu-8906c1644664

Single machine, multiple GPUs working on separate tasks

First of all, it’s not about SLI, you may heard about in the good old days of computer gaming. SLI is a technology to link 2–4 GPUs to share the work on rendering an image. It’s only about rendering graphics.

In CUDA (that is about computations, not graphics) you can directly access any available GPU in your system, so just add several GPUs and use any of it. You can write your program to do anything you want, loading data into any GPU and running computations on a GPU of choice.

Usually deep learning engineers do not write CUDA code, they just use frameworks they like (TensorFlow, PyTorch, Caffe, …). In any of these frameworks you can tell the system which GPU to use.