Rhino Render not using quadro card

I recently purchased a thinkpad p52 workstation with a quadro 2000 card for design class and when I go to hit render, rhino is maxing out the xeon processor, but the quadro is left untouched. I set rhinocycles_select device to “1” which was listed as the cuda device and it’s still not using it. The nvidia drivers are up to date and I selected the workstation 3d as per instructions in nvidia’s control panel but still no go. Does the built in rhino render not use a gpu? Is there anything I can do to stop these simple renders from taking 20 minutes?

When you use _Render you’ll be using the old-style renderer, which is not using the GPU.

You’ll need to use the Raytraced viewport mode to utilize your GPU. To get render results from the Raytraced viewport you’d use the _ViewCaptureToClipboard and _ViewCaptureToFile commands. If you are working with more raytrace-intensive scenes (size-wise, or complexity otherwise), you might want to consider using _TestPackageManager and install the rhino-render-next package, which gives you _Render capability with the GPU using the Rhino Render Next renderer (which is Raytraced, but then for _Render).

So I follow the instructions above and at best got the card to 30% utilization but it still took a long time (1 1/2 hours) to render a 1k resolution scene. The xeon was maxed out during the whole rendering process though. I totally get that pretty renders aren’t important for actual design work but…they’re pretty. Is this time pretty typical for a scene that was simple (20 odd cylinders with a texture)?

Did you also change your current renderer to Rhino Render Next, and ensured your Quadro is selected in the Tools > Options > Cycles settings page (under CUDA)?

If you upload your test 3dm file here and post your system info so folks can get a better idea what you are talking about you’ll probably get some pretty good answers.

1 Like