Cycles differences between CPU and GPU in this World nodes #53348

New Issue

YAFU · 2017-11-19T14:51:08+01:00

YAFU commented

2017-11-19 14:51:08 +01:00

System Information
Kubuntu 14.04 - GTX 960

Blender Version
Broken: 2.79 official, master

Short description of error
First of all, I'm not good with nodes. This may be due to a bug, or a bad configuration of nodes on my part. If it is the latter, sorry for inconveniences.
There are differences between GPU or CPU. In 2.79 release CPU render some circular stripes, still result different from GPU. From Master CPU does not render any texture.

Exact steps for others to reproduce the error

World_cpu_gpu_differences.blend
Open scene. Render with CPU or GPU (or combined from master)

**System Information** Kubuntu 14.04 - GTX 960 **Blender Version** Broken: 2.79 official, master **Short description of error** First of all, I'm not good with nodes. This may be due to a bug, or a bad configuration of nodes on my part. If it is the latter, sorry for inconveniences. There are differences between GPU or CPU. In 2.79 release CPU render some circular stripes, still result different from GPU. From Master CPU does not render any texture. **Exact steps for others to reproduce the error** [World_cpu_gpu_differences.blend](https://archive.blender.org/developer/F1159561/World_cpu_gpu_differences.blend) Open scene. Render with CPU or GPU (or combined from master)

YAFU commented

2017-11-19 14:51:08 +01:00

Changed status to: 'Open'

YAFU commented

2017-11-19 14:51:08 +01:00

Added subscriber: @YAFU

Sergey Sharybin commented

2017-11-22 17:57:08 +01:00

Added subscriber: @Sergey

Sergey Sharybin self-assigned this 2017-11-22 17:57:08 +01:00

Sergey Sharybin commented

2017-11-22 17:57:08 +01:00

Root of the issue is coming from precision issue on GPU. The thing here is that gradient texture always gives black color is such configuration, which effectively causes division by zero. Now, even though we use safe division (as in, we check whether denominator is zero or not), this does not help on GPU. Root of this issue goes to --fast-math we use for CUDA. Disabling fast math solves this issue, but gives about 15% of slowdown.

Can try using some more granular flags, but this requires some time.

Root of the issue is coming from precision issue on GPU. The thing here is that gradient texture always gives black color is such configuration, which effectively causes division by zero. Now, even though we use safe division (as in, we check whether denominator is zero or not), this does not help on GPU. Root of this issue goes to `--fast-math` we use for CUDA. Disabling fast math solves this issue, but gives about 15% of slowdown. Can try using some more granular flags, but this requires some time.

Sergey Sharybin removed their assignment 2017-11-23 10:25:15 +01:00

Brecht Van Lommel was assigned by Sergey Sharybin

2017-11-23 10:25:15 +01:00

Sergey Sharybin commented

2017-11-23 10:25:15 +01:00

Added subscriber: @brecht

Sergey Sharybin commented

2017-11-23 10:25:15 +01:00

@brecht. didn't manage to find combination of flags which will solve the artifacts without introducing speed regression. More and more tempting to declare "just avoid numerical issues in shaders". But maybe you want to have a try here? (maybe disable fast divisions and use approx intrinsic when it's needed?)

YAFU commented

2017-11-23 12:12:51 +01:00

Hi.
I was just trying to get a quick background for the BA weekend challenge. I'm not even sure if those nodes are being used well here, and neither is it common to use it in World nodes. If you ask me, I prefer speed :) . But I have no idea, perhaps the problem that jumped here could affect other cases, and since we have GPU + CPU combined render, differences are more sensitive.
The nodes in Material render similar on GPU and CPU:
World_cpu_gpu_differences_1.blend

Hi. I was just trying to get a quick background for the BA weekend challenge. I'm not even sure if those nodes are being used well here, and neither is it common to use it in World nodes. If you ask me, I prefer speed :) . But I have no idea, perhaps the problem that jumped here could affect other cases, and since we have GPU + CPU combined render, differences are more sensitive. The nodes in Material render similar on GPU and CPU: [World_cpu_gpu_differences_1.blend](https://archive.blender.org/developer/F1178974/World_cpu_gpu_differences_1.blend)

blender-admin commented

2017-11-23 16:10:38 +01:00

This issue was referenced by blender/cycles@dbe9eec374

This issue was referenced by blender/cycles@dbe9eec374916d07a3389068891d2fffb2921823

blender-admin commented

2017-11-23 16:10:38 +01:00

This issue was referenced by e50ed90e4d

This issue was referenced by e50ed90e4dff3a01fd5e3211f3571a34e1d6855d

Brecht Van Lommel commented

2017-11-23 16:41:00 +01:00

In the gradient texture we end up computing 1 - length(unit_length_vector) which is supposed to be exactly zero, but of course in practice float precision means it's not. I'll bias it a little bit now to make it exactly zero. We have tweaks like that in other places too, like the checker texture. It helps to make things a bit more robust also when rendering with only the CPU or GPU.

In general I think we want to keep the fast math enabled, even if it risks issues like this.

In the gradient texture we end up computing `1 - length(unit_length_vector)` which is supposed to be exactly zero, but of course in practice float precision means it's not. I'll bias it a little bit now to make it exactly zero. We have tweaks like that in other places too, like the checker texture. It helps to make things a bit more robust also when rendering with only the CPU or GPU. In general I think we want to keep the fast math enabled, even if it risks issues like this.

Brecht Van Lommel commented

2017-11-23 17:41:05 +01:00

Changed status from 'Open' to: 'Resolved'

Brecht Van Lommel closed this issue

2017-11-23 17:41:05 +01:00

Steve Hargreaves commented

2017-12-02 15:37:02 +01:00

Added subscriber: @Roken

Steve Hargreaves commented

2017-12-02 15:37:02 +01:00

poolproblem.blend This exhibits the same problem. The experimental GPUCPU render really highlights it.

[poolproblem.blend](https://archive.blender.org/developer/F1264212/poolproblem.blend) This exhibits the same problem. The experimental GPUCPU render really highlights it.

Sign in to join this conversation.

No Label

Download

What's New

Blender Studio

Manual

Developers Blog

Documentation

Benchmark

Blender Conference

Development Fund

One-time Donations

Cycles differences between CPU and GPU in this World nodes #53348