Gigapixel 8 Redefine above 3 creativity and other settings = it stucks and computer isnt render anymore

Hello,
I have the following problem that Gigapixel 8 takes ages to render even when it is set to x 1 (upscale). So a preview is not possible at first, when I click on Export Image, Gigapixel 8 turns a 200 KB file into a 200 MB file. I have already waited over 2 hours and nothing has happened.

If I leave creativity on 1 or 2, then my PC hangs = it renders, after 5 minutes I’m done. As soon as I change other settings, such as creativity to 3, I can start the render process, but my computer stops working. Because everything runs smoothly… on creativy 1 or 2, my computer is lagging (but its ok for me…) but I want test the creativity on 6. But the render process wont start.

So one indicator is when everything hangs = the render process is in progress…

My computer:
Ryzen 7 5700X
32 GB RAM
Radeon RX 6800 XT

gigapixel 8 is already on the newst update / patch.

What is the problem?!

Screenshot on the rendering process, no cpu/gpu usage…
on creativity 3 or above.

creativity settings on 1 or 2 - high usage of gpu…:

now with video

the first step: creativity on 1 = it start to rendering.
the other step: creativity on 3 = it stucks.

1 Like

I have the exact same issue. Anything over creativity level 2 doesn’t begin to render at all. The rendering dialog shows, but the application/PC does nothing. I am also using an AMD GPU (6900XT). I’m wondering if that has something to do with it.

They can’t help, they are just saying, you need to update the gpu driver, but I have done this, and still it doesn’t work… I can’t use this program, even the cloud service doesnt work. They have said, I should reinstall windows, I’ve done, but it doesn’t work.

I don’t think it’s drivers. I’m using the latest driver and it’s less than two weeks old. I think it’s lack of CUDA support on AMD cards. I checked the logs and it looks like 3 and above are using some different models than when I run it on 1 or 2. It’s possible those models require CUDA? I’m not really sure. onnxruntime should be able to use DirectML as well as CUDA, but maybe not for all models. I tried updating the DirectML libraries, but that didn’t help. I also tried updating the onnxruntime libraries, but I think Gigapixel is using a modified version (the file size is larger eventhough it’s based on an older version) and Gigapixel won’t launch on the latest version.

This is the error it’s throwing. After that it stops all workers and basically just sits there until you tell it to stop.

[2025-02-03 07:57:48.085, 0.00 μs] [29e0] Info | [AIE] ONNX error at 256: Non-zero status code returned while running DmlFusedNode_0_33 node. Name:‘DmlFusedNode_0_33’ Status Message: D:\a_work\1\s\onnxruntime\core\providers\dml\DmlExecutionProvider\src\DmlGraphFusionHelper.cpp(1066)\onnxruntime.dll!00007FFD97B6FE31: (caller: 00007FFD97BFD725) Exception(4) tid(29e0) 887A0001 The application made a call that is invalid. Either the parameters of the call or the state of some object was incorrect.