LuxMark - The OpenCL CPU&GPU benchmark

Printable View

02-27-2011, 02:48 PM
fellix_bg

LuxMark - The OpenCL CPU&GPU benchmark

LuxMark is an OpenCL benchmark based on LuxRender. LuxRender is a physically based and unbiased rendering engine. Based on state of the art algorithms, LuxRender simulates the flow of light according to physical equations, thus producing realistic images of photographic quality - Gallery.

Info and Download

The default (start-up) benchmark mode is the GPU path.

Here is the my GTX570 (825/2100MHz):

http://img222.imageshack.us/img222/6468/99166196.jpg

Happy benchmarking!
02-27-2011, 03:42 PM
Toysoldier

GTX580 (900/2004MHz)

http://i132.photobucket.com/albums/q...er/LuxMark.png
02-27-2011, 03:51 PM
Toysoldier

GTX580 (772/2004Mhz - default speed)

http://i132.photobucket.com/albums/q.../LuxMark_1.png
02-27-2011, 04:03 PM
Toysoldier

GTX580 (825/2100Mhz)

http://i132.photobucket.com/albums/q...ark8252100.png
02-27-2011, 04:19 PM
Sam_oslo

Here is my undervalued (0.987v) 24/7 setup

GTX560 (910/2400MHz):
http://img.techpowerup.org/110227/Capture028.jpg
02-28-2011, 12:51 AM
NapalmV5

900/2200 - 100mhz pcie
http://i51.tinypic.com/345ouvs.png
02-28-2011, 02:08 AM
kaktus1907

@NapalmV5 sweet mother of mercy :shocked: :shocked:

OpenCL GPU Only = 5092
GTX 460 1GB @950/1900/4600MHz
http://www.abload.de/thumb/desktop_2...25_19_6m0a.png

Native CPU = 2393
Q9650 @4.00GHz
http://www.abload.de/thumb/desktop_2...25_19_gmbr.png
02-28-2011, 05:18 AM
zalbard

2x 6970

http://i301.photobucket.com/albums/n...rd/LuxMark.jpg
02-28-2011, 11:46 AM
bamtan2

very nice, very interesting. luxmark may have practical implications, unlike some synthetic benchmarks.
02-28-2011, 12:45 PM
freeloader

And bringing up the bottom end just for fun.....AMD C50 1GZ dual core laptop.

http://img684.imageshack.us/i/luxmark.jpg/
02-28-2011, 02:02 PM
Roadhog

Almost hit 20k on 3, 6970's @ 950/1450.

http://i.imgur.com/zYjXE.jpg
02-28-2011, 03:30 PM
STEvil

GTX480 @ 800/2000 - 3800
E5450's @ 3.6ghz (8 cores) CPU Native - 3450

WinXP-64

Doesnt want to run on my Win7 system with the westmere's and 4870x2..
03-01-2011, 01:04 AM
Dade

Quote:

Originally Posted by NapalmV5

900/2200
30709 LuxMarks

Hi NapalmV5, I'm one of the LuxMark developer. Your score is the highest I have seen up to now. Cheers, you have redefined my idea of BFG9000 :D
03-01-2011, 02:57 AM
STEvil

hey dade any reason why the gtx480 gets half the score of a gtx570? does OS make that much difference, or could it be driver version?
03-01-2011, 03:37 AM
Dade

Quote:

Originally Posted by STEvil

hey dade any reason why the gtx480 gets half the score of a gtx570? does OS make that much difference, or could it be driver version?

Are you using the latest drivers ? NVIDIA has highly improved their OpenCL support in the latest release.

According Anandtech (http://www.anandtech.com/show/4135/n...-250-market/15), a 480gtx should score about the same value of a 570gtx with latest drivers.

If you check older Anandtech's reviews (http://www.anandtech.com/show/4061/a...eon-hd-6950/23), you can see like NVIDIA as nearly doubled the scores with latest drivers.

AMD has also improved the performance of HD6xxx family of a good 15-30% in the latest driver release.
03-01-2011, 05:06 AM
STEvil

Thats exactly what it was I guess.

Went from 3800 with 260.69's to 8500 with 267.24's at 800/2000
03-01-2011, 11:29 AM
NapalmV5

Quote:

Originally Posted by kaktus1907

@NapalmV5 sweet mother of mercy :shocked: :shocked:

Quote:

Originally Posted by Dade

Hi NapalmV5, I'm one of the LuxMark developer. Your score is the highest I have seen up to now. Cheers, you have redefined my idea of BFG9000 :D

thanks guys :toast: score goes to show how well youve implemented multi gpu support :up:

950/2200 - 100mhz pcie
http://i52.tinypic.com/n3n6zp.png
03-01-2011, 01:50 PM
jogshy

After taking a look at the code, I wonder how dependent is this test on the PCI express. I've been never a big fan of passing tons of data by the bus.
Any of you can make a quick test pls? My BIOS does not allow to change the PCI multiplier... but I bet if you put a monster GPU and you put it to 2x a modest GPU running at 16x might be faster.

Also, I think the automatically assigned work size of 64 thread/block may cause an slowdown.

My ... strange.... results:

OpenCL - GPUs only ( a Radeon 5750, catalyst 11.2 ) -------------> 1784 marks
http://img602.imageshack.us/img602/7909/test1ww.jpg
http://img171.imageshack.us/img171/8533/test2e.jpg
Native CPU ( an obsolete i7 920 ) ----> 3342 marks

So, in my case, the CPU is 2x faster than the GPU... Shouldn't be a GPU faster than a CPU rendering? That's the whole point of Octane, VRayRT and other GPGPU renderers after all...
03-01-2011, 11:21 PM
Brahmzy

This is an awesome Benchmark - very useful.
I'll post some screens later today..
03-02-2011, 01:37 AM
Dade

Quote:

Originally Posted by jogshy

After taking a look at the code, I wonder how dependent is this test on the PCI express. I've been never a big fan of passing tons of data by the bus.
Any of you can make a quick test pls? My BIOS does not allow to change the PCI multiplier... but I bet if you put a monster GPU and you put it to 2x a modest GPU running at 16x might be faster.

Every 5 seconds, the images rendered by each GPU are downloaded to the CPU memory to be merged a drawn on the screen. It is a very little amount of data for the bandwidth of PCIe bus.
I have 2x5870 plus a 5850 installed on a 4x PCIe slot. I have done some test, moving the 5850 between 16x slots and 4x slot: I haven't experienced any performance difference.

Jogshy, anyway we need a feature like GPU-to-GPU direct transfer announced for CUDA 4.0 to solve this problem (if it becomes a problem with many GPUs).

Quote:

Originally Posted by jogshy

Also, I think the automatically assigned work size of 64 thread/block may cause an slowdown.

A little trick: you can change that value by editing the file scenes/luxball/render-hdr.cfg. Just edit the following line:

opencl.gpu.workgroup.size = 64

Quote:

Originally Posted by jogshy

So, in my case, the CPU is 2x faster than the GPU... Shouldn't be a GPU faster than a CPU rendering?

i7 920 is more expansive that a HD5750 and the GPU is still faster. I don't read it as a bad result for the CPUs Vs GPUs "holy war".

Quote:

Originally Posted by jogshy

That's the whole point of Octane, VRayRT and other GPGPU renderers after all...

Jogshy, I assume you are the same of ompf.org forums; you know that there have been a lot of discussion about this point. Many people believe there is a lot of marketing vaporware in CPUs Vs GPUs "holy war".

My opinion is somewhat in the middle, GPUs computing offers great value (especially for home/office users) but it is no where near the 100x faster promised by the marketing.
03-02-2011, 02:36 AM
Toysoldier

I tried changing the work group size to the suggested value of 640 and it throws an error
OpenCL ERROR: clEnqueueNDRangeKernel(-54)

So I tried with different work group sizes :

32 : OK ~ 7920
64 : OK ~ 8780
128: OK ~ 8650
256: OK ~ 8412
512: OK ~ 7580
640: OpenCL ERROR: clEnqueueNDRangeKernel(-54)
768: OpenCL ERROR: clEnqueueNDRangeKernel(-54)
1024: Rendering thread ERROR: clSetKernelArg(-38)

So it looks like 64 is the best workgroup size, at least on a GTX580 with the current codepath.
03-02-2011, 08:14 AM
jogshy

Quote:

Originally Posted by Dade

i7 920 is more expansive that a HD5750 and the GPU is still faster. I don't read it as a bad result for the CPUs Vs GPUs "holy war".

Faster? My 5750 is actually 2x slower than the CPU as you can see in the screenshots.
Ok, the i920 is 2x more expensive than a 5750 so, relatively, the ATI GPU==CPU. A bit dissapointing.

Quote:

Originally Posted by Dade

Every 5 seconds, the images rendered by each GPU are downloaded to the CPU memory to be merged a drawn on the screen. It is a very little amount of data for the bandwidth of PCIe bus.

Taking a look to your code seems you're generating the rays in the CPU, then passing then to the GPU to perform the ray tests and then you bring back the results to the CPU. That implies a lot of PCI transfers, latency and synchronization. Can anybody with a decent BIOS test how much it scales, pls? My BIOS sux :(
03-02-2011, 08:37 AM
kaktus1907

pcie bandwith numbers
16x pcie 2.0

Idle
http://www.abload.de/img/desktop_2011_03_02_18_x7uf.png

Load
http://www.abload.de/img/desktop_2011_03_02_18_f7pk.png
03-02-2011, 09:31 AM
jogshy

Quote:

Originally Posted by kaktus1907

pcie bandwith numbers

Nak, can your BIOS change the PCI multiplier?
03-02-2011, 09:55 AM
soya_crack

Hi guys,

Do you know if a SLI Bridge is of advantage here?

@Napalm

Did you use one?

I am hitting 15k on 3x 560 GTX (not overclocked).
03-02-2011, 11:53 AM
NapalmV5

@soya_crack

yes sli bridge and 100mhz pcie
03-02-2011, 01:51 PM
Dade

Quote:

Originally Posted by jogshy

Taking a look to your code seems you're generating the rays in the CPU, then passing then to the GPU to perform the ray tests and then you bring back the results to the CPU. That implies a lot of PCI transfers, latency and synchronization.

No, that is the hybrid rendering mode, look for PathGPU rendering mode (the sources are in the pathgpu directory: http://src.luxrender.net/luxrays/fil...uxmark/pathgpu). It is the one used in LuxMark.
03-02-2011, 03:57 PM
Brahmzy

2 x GTX570 SLi (Default 732/1900)
http://img199.imageshack.us/img199/7863/capture2er.jpg

2 x GTX570 SLi (850/2000)
http://img843.imageshack.us/img843/4927/capture1gr.jpg
04-19-2011, 10:35 PM
NapalmV5

and heres sli @ 950/2300 100mhz pcie

http://img98.imageshack.us/img98/8420/unledzt.png
04-20-2011, 09:04 AM
-=DVS=-

http://i.imgur.com/lmtW9.jpg
09-04-2011, 07:33 AM
SUPERKAMES

1 Attachment(s)

2* HD6950 unlocked @ 900-920 mhz .
09-04-2011, 07:45 AM
kaktus1907

^^

wow AMD's opencl drivers deliver amazing perf improvement.. almost on par with a GTX 580
09-20-2011, 04:29 PM
mattkosem

1 Attachment(s)

Here's what I get on an A8 Llano at 23x152 (just CPU-Z since GPU-Z doesn't show anything helpful on this chip).

--Matt
10-02-2011, 01:09 PM
Gilgamesh

1 Attachment(s)

Good OpenCl improvement from Catalyst 11.9 :)

OLD 11.6 bench 2xHD6950@6970 - 935/1400 - 18182
http://www.pctunerup.com/up/results/...k_Capture2.JPG

NEW 11.9 bench 2xHD6950@6970 - 935/1400 - 22352
http://www.pctunerup.com/up/results/...6_luxmark3.JPG

@ 960/1400 - 22843

Attachment 120744