GPU Usage Modes

High-end Quadro GPUs in workstations can operated in different usage, or processing, modes. By setting specific usage modes for each GPU, you can control how much processing power is dedicated to graphics processing and how much is dedicated to computational processing.


NVIDIA Quadro Products

Dedicate to Graphics needs: In this mode, the GPU's resources are used for graphics processing.

Use for Graphics and compute needs (Enable CUDA compute): In this mode, GPU resources are allocated for both graphics as well as compute tasks.


NVIDIA Tesla C2075 or Tesla K20

These Tesla GPUs are dedicated to compute needs, meaning that 100% of the GPU's resources are used for compute tasks. You cannot use these Tesla cards to drive a display.


Related topics