LuxrenderGPU

[KyungSoo]

8 GPU's machine?
Just because I am curious, could you give me details about your configuration?
I mean, Is it two 4-Way sli machines, or 4 machines each double SLI?

It is one machine which has 4 GTX-295(dual GPU) graphics cards.
And I did not set SLI settings.

[fredmj]

Thanks for answer. Yep I didn't remember GT 295 was a double GPU card. Well you should reach the best performances. And what is your double CPU mother board? omething like a Supermicro(c) one?

[KyungSoo]

Thanks for answer. Yep I didn't remember GT 295 was a double GPU card. Well you should reach the best performances. And what is your double CPU mother board? omething like a Supermicro(c) one?

Yes it is a Supermicro, 1U GPU server, which has two PCIe slots.

[Dade]

And SmallluxGPU v1.3Beta1 test was not successful.
I met following messages.
[Device::GeForce GTX 295] RayIntersection thread started
[Device::GeForce GTX 295] RayIntersection ERROR: clEnqueueWriteBuffer(-5)

You need newer NIVIDA driver to run v1.3beta1. Something in the range of 195+, NVIDIA drivers were quite buggy before 195.xx release.

[KyungSoo]

And SmallluxGPU v1.3Beta1 test was not successful.
I met following messages.
[Device::GeForce GTX 295] RayIntersection thread started
[Device::GeForce GTX 295] RayIntersection ERROR: clEnqueueWriteBuffer(-5)

You need newer NIVIDA driver to run v1.3beta1. Something in the range of 195+, NVIDIA drivers were quite buggy before 195.xx release.

Thank you for your information, new nVidia driver makes it works.
And there is large speed up, too, from 2,500K to 3,400K.
But, the load of GPUs are not improved, yet. They are still around 40%s.

[Dade]

Thank you for your information, new nVidia driver makes it works.
And there is large speed up, too, from 2,500K to 3,400K.
But, the load of GPUs are not improved, yet. They are still around 40%s.

Quite impressive to see 8 GPUs at work. I have a new SmallLuxGPU version where you can assign even multiple threads to keep a GPU busy: all GPUs (i.e. intersection devices) are grouped together under a virtual device seen by the system like a single GPU (the workload is dynamically assigned to the less busy real GPU). Than I can define any number of threads to produce work for this single virtual device (for instance you can have 4 threads producing work for 2 GPUs).

However it would not solve the problem in your case because I assume you have more GPUs than CPU cores on your test system. I'm afraid the only option in a system like your is to make CPU code run faster and/or translate more work from the CPU to the GPUs.

There is also the option to increase the complexity of the scene, for instance you could try the scene with 2.7M triangles (but probably you need a more complex scene): http://davibu.interfree.it/opencl/small ... scenes.tgz

[KyungSoo]

Thank you for your information, new nVidia driver makes it works.
And there is large speed up, too, from 2,500K to 3,400K.
But, the load of GPUs are not improved, yet. They are still around 40%s.

Quite impressive to see 8 GPUs at work. I have a new SmallLuxGPU version where you can assign even multiple threads to keep a GPU busy: all GPUs (i.e. intersection devices) are grouped together under a virtual device seen by the system like a single GPU (the workload is dynamically assigned to the less busy real GPU). Than I can define any number of threads to produce work for this single virtual device (for instance you can have 4 threads producing work for 2 GPUs).

However it would not solve the problem in your case because I assume you have more GPUs than CPU cores on your test system. I'm afraid the only option in a system like your is to make CPU code run faster and/or translate more work from the CPU to the GPUs.

There is also the option to increase the complexity of the scene, for instance you could try the scene with 2.7M triangles (but probably you need a more complex scene): http://davibu.interfree.it/opencl/small ... scenes.tgz

Sean with 2.7M triangles did not increase the speed, as screen dump.
I guess, you should move the ray generation codes from CPU to GPUs as you mentioned,
"I'm afraid the only option in a system like your is to make CPU code run faster and/or translate more work from the CPU to the GPUs."
That would increase the speed of CPU/GPU threads, in general (one or two GPUs) cases, too.

[vertigo262]

This might sound like a stupid question and off topic, but I'm trying to load SmallLuxGPU on a laptop with an ATI radeon mobility x1400
and it errors saying I don't have opencl installed.

Is the opencl for the ati in the catalyst driver install, or is it a seperate install that I need from ati?

any input would be greatly appreciated

[Dade]

This might sound like a stupid question and off topic, but I'm trying to load SmallLuxGPU on a laptop with an ATI radeon mobility x1400
and it errors saying I don't have opencl installed.

Is the opencl for the ati in the catalyst driver install, or is it a seperate install that I need from ati?

any input would be greatly appreciated

You need to install also the ATI Stream SDK (http://developer.amd.com/GPU/ATISTREAMS ... fault.aspx): OpenCL.dll is not yet included in main Catalyst release. However, check the list of supported GPUs, it doesn't look like the x1400 is in the list.

[SATtva]

OpenCL is supported only in HD series cards and above. Unfortunately, X1400/X1600 are already much outdated (and even incompatible with the latest 9.x drivers).