Rhino 5 not using GPU in viewports?

Hi everyone !

I’ve been using and loving Rhino for a few years now, but this is my very first post in the forum !

I recently built a new workstation to work mostly in Rhino.
I’m actually working on quite a large 3D model of a part of NYC.
The problem is that the display is veeeery slow while rotating, zooming, panning, even in Wireframe Mode.

So I used HW Monitor to see the usage of my CPU and my GPU (GTX 1080) and unfortunately, It appeared that when i’m moving inside of a viewport, there is no effect on the GPU, as if it was NOT used at all, whereas the CPU usage is 100%.
There is also something else very strange : when I check the OpenGL properties, Rhino seems to detect only 4GB of VRAM, when the GTX 1080 has 8GB.

I bought this quite expensive video card because I thought I would get a big increase in Rhino (viewport) performance, and I’m now quite disappointed.

The GTX1080 is working fine with Thea render at least.

Please could someone try to help me ?

I’m using the latest version of Rhino 5 (SR 13) and the latest Nvidia drivers.
Hardware acceleration is ON.
Here are the details of my system :

Windows 10 (64 bits)
Intel I7 6850K (6 cores at 3.6 Ghz)
32 GB of DDR 4 RAM (2400 Mhz)
GTX 1080 (8GB VRAM)
Samsung M.2 SSD

I shifted your post to the Hardware sub category where it will be seen by people better able to help.

Oh I’m sorry, I didn’t pay attention.
Thank you !

What does the Holomark utility tell you?

Here are the results of the test :

Holomark 2 v2,2,03

Total Score: 41407
Total Runtime: 152.47 sec

GPU scores: 24290
GPU_01 - 400.00 fps - Cube 4 tests
GPU_02 - 49.50 fps - UDT Shape
GPU_03 - 85.50 fps - Wireframe
GPU_04 - 58.10 fps - Shaded
GPU_05 - 40.80 fps - Rendered
GPU_06 - 39.50 fps - Block Rendered
GPU_07 - 22 units Nurbs @ 5 fps in Wireframe
GPU_08 - 14 units Nurbs @ 5 fps in Shaded
GPU_09 - 50 units Nurbs @ 5 fps in RenderSpeed
GPU_10 - 128.20 fps - RenderMesh Render
GPU_11 - 156.30 fps - RenderMesh RenderSpeed
GPU_12 - 129.90 fps - JoinedMesh Render
GPU_13 - 149.30 fps - JoinedMesh RenderSpeed
GPU_14 - 5 units mesh @ 15 fps in Shaded
GPU_15 - 8 units mesh @ 15 fps in Render
GPU_16 - 26 units mesh @ 15 fps in RenderSpeed
GPU_17 - 113.60 fps - mesh in Rendered Studio
GPU_18 - 14.80 fps - Nurbs in Rendered Studio
GPU_19 - 29.50 fps - Block Illustration
GPU_20 - 87.00 fps - 2D single
GPU_21 - 9.50 fps - 2D massive (20x)

CPU scores: 17117
CPU_01 - 8.55 sec - Booleans and Contours
CPU_02 - 1.06 sec - Twist and Taper (UDT)
CPU_03 - 3.91 sec - Meshing Mini
CPU_04 - 0.03 sec - Extract Render Mesh
CPU_05 - 0.04 sec - Join Render Mesh
CPU_06 - 11.03 sec - Reduce Mesh
CPU_07 - 1.57 sec - Calculating Technical display
CPU_08 - 3.02 sec - Making Silhouettes


NVIDIA GeForce GTX 1080 - 4095.0 MB

Intel® Core™ i7-6850K CPU @ 3.60GHz
NumberOfCores: 6 NumberOfLogicalProcessors: 12
MaxClockSpeed: 3.6 GHz

TotalPhysicalMemory: 32.0 GB

Microsoft Windows 10 Professionnel

  • None - 64 bits

Rhino 5 sr 13 64 bit

AT a quick glance, that seems in-line with other similarly configured systems.

Have your tried the V6 WIP?
The same test won’t run in V6 but V6 uses a new updated display system that should be a lot faster.

How big is your model? File size, # of objects, # of blocks?

Hopefully someone with a lot of video card experience will chime in about why Rhino only finds 4 out of 8GB on the 1080.

Rhino tries its best to harvest information about the graphics system and sometimes gets it wrong. This information is only used as exactly that - information. In no way is this used to limit its performance.

Interesting, I never noticed the incorrect RAM display.
BTW, V6 WIP displays it correct.

Not sure which Rhino is right about the OpenGL version:


Thank you all for your quick answers !

I tried on the WIP and it’s definetly working way better with my gpu !
But unfortunately, it’s difficult for me to work in V6, as I need all my plugins and stability.

So I guess there is no real workaround in V5 and I will have to wait until the release of V6.
Too bad… I hope it’s going out soon.

Thanks everyone !