hexagon logo

How can I use my NVIDIA P1000 as a GPU to accellerate the simulations?

I would like to use the graphic card on my workstation (NVIDIA P1000) as a GPU. There is no monitor connected to the card as I'm accessing it via remote. So, I thought it should be possible. The P1000 is CUDA capable and I've also installed the CUDA Toolkit as per the Marc User Manual. But when I run the solver with the gpgpu command it still runs only on the CPU. I'm using Marc 2018.1.0.
Does anyone has experience using GPU and can give me a hint how to go forward?