[Bf-cycles] CUDA Toolkit 6.5 and Volume on GPU

Thomas Dinges blender at dingto.org
Fri Aug 22 01:20:09 CEST 2014


Heads up regarding memory usage (tested with GPU-Z):
Vanilla Master with CUDA 6.0 and 6.5 are roughly the same (69 vs. 71 
Megabyte for a default cube).

With Volume (__VOLUME__ and __VOLUME_SCATTER) enabled the kernel needs 
additional 15 Megabyte (85 Megabyte for the default cube). I think that 
is a reasonable price for Volume support on GPU.

@Zanqdo: When compiling with SSS we still have a much larger memory 
footprint, 148 Megabyte in a quick test, don't think we can enable this 
soon.

@all
I can only test sm_21. sm_20, 30, 35 and 50 will be interesting too. If 
you compile yourself, you could help me here.

Test Render Time and Memory Usage with CUDA Toolkit 6.5.
1) Vanilla
2) __VOLUME__ enabled
3) __VOLUME__ and __VOLUME_SCATTER__ enabled.

These 2 options can be enabled in cycles/kernel/kernel_types.h, lines 80/81.

Best regards,
Thomas

Am 21.08.2014 um 23:54 schrieb Thomas Dinges:
> Hi everyone,
> NVIDIA just released the CUDA Toolkit 6.5:
> https://developer.nvidia.com/cuda-toolkit
>
> I recommend an update for 2 reasons:
> 1) It improves rendering performance.
> 2) It now supports Visual Studio 2013, so no other VS or SDK
> installations are needed anymore on Windows.
>
> I ran some benchmarks on Windows 7 x64, with a Geforce 540M (sm_21),
> master 2e5e1e2.
> - cornell_box.blend (256x256 Tiles, 100 Samples)
> - bmw.blend (128x128 Tiles, 100 Samples)
>
> 1) Master (Vanilla) with Toolkit 6.0
> - cornell: 57.47s
> - bmw: 1:53.63min
>
> 2) Master (Vanilla) with Toolkit 6.5
> - cornell: 54.10s
> - bmw: 1:50.10min
>
> 3) Master (with __VOLUME__) with Toolkit 6.5
> - cornell: 56.15s
> - bmw: 1:52.51min
>
> 4) Master (with __VOLUME__ and __VOLUME_SCATTER__) with Toolkit 6.5
> - cornell: 56.99s
> - bmw: 1:54.37min
>
> I think we have 2 choices after the CUDA Toolkit upgrade:
> 1) Keep feature set as is, benefit from better performance
> 2) Enable Volume (and Volume Scatter) on GPU, performance stays roughly
> the same
>
> Best regards,
> Thomas
>
>
>
>
> _______________________________________________
> Bf-cycles mailing list
> Bf-cycles at blender.org
> http://lists.blender.org/mailman/listinfo/bf-cycles



More information about the Bf-cycles mailing list