Multiple gpu for tensor flow

[kmatt]

Hi everyone, wondering if anyone has done this and how it went for you. I currently have an RX580 in my rig and want to use an extra gtx 970 I've got for it's cuda cores to work with tensorflow. I'm wondering if I can just install the 970 in my second pci-e slot and get drivers for it without using it for my main display. Tensorflow doesn't play well with amd cards from what I've been able to figure out so being able to just use the 970 would be a lot easier for me than forcing the amd card to work through (what I think is called) open cv. 

 

Any advice or knowledge would be appreciated. Wish me luck as I jump into AI and deep learning as a noob

 

[Electronics Wizardy]

Yea you can just have 2 gpus and it will just work asuming your board is compatible. You can easily install gpu drivers from different manafactures.

[kmatt]

27 minutes ago, Electronics Wizardy said:

Yea you can just have 2 gpus and it will just work asuming your board is compatible. You can easily install gpu drivers from different manafactures.

Thank you!