Division of graphical power between Internal GPU and RTX 3070 GPU

Dear Nvidia Community.

I have an RTX 3070 (ASUS 90YV0FAC ASUS ROG-STRIX-RTX3080-O12G-GAMING) and I have a built in GPU as well in my Desktop PC that isn’t RTX. I am performing computationally extensive tasks. Is it possible to do the backend calculation on the RTX 3070 GPU and the visulaization on the internal GPU.

I’m looking forward to a kind response.

Best Regards,
Abdul Muqeet Waheed
Robotics Engineer

Hello Abdul!

Of course this is possible. Depending on your operating system it might already allow you to choose dedicated GPUs for certain tasks.

But if you are running compute tasks then you will most likely work with CUDA which offers direct control over which GPU to use for what tasks. If you use CUDA you should familiarize yourself with the CUDA documentation. Look for chapter 6.1 Device Management.

If you are using other tools, like for example TensorFlow, they also have specific commands to choose a certain GPU, like tf.device().

I hope this helps!