Benchmarking my cuda cloth

Fadingz · March 1st, 2012, 19:17

Hello all, here is my cuda cloth simulator

Please test out the three resolutions and let me know the performance.
I will need your video card and fps.
Thank you!

There are nxn particles with nxnx64 springs.
So,
1) 66k particles, 4M springs
2) 262k particles, 17M springs
3) 1M particles, 64M springs

Gravity always points down.
Drag your mouse around to rotate the view.

NOTE: YOU HAVE to have a cuda supported NVidia card!

Please use 1_0_ if you know your card supports cuda, but not which version.
If you know your card supports cuda compute 2.0 and above, please use 2_0_

Here is a chart of gpus and cuda compute versions:

Code:

CUDA-Enabled Tesla GPU Computing Products
Tesla
Tesla Workstation Products GPU     Compute Capability
Tesla C2075     2.0
Tesla C2050/C2070     2.0
Tesla C1060     1.3
Tesla C870     1.0
Tesla D870     1.0
Tesla Data Center Products GPU     Compute Capability
Tesla M2050/M2070/M2075/M2090     2.0
Tesla S1070     1.3
Tesla M1060     1.3
Tesla S870     1.0
CUDA-Enabled Quadro Products
Quadro
Quadro Desktop Products GPU     Compute Capability
Quadro 6000     2.0
Quadro 5000     2.0
Quadro 4000     2.0
Quadro 4000 for Mac     2.0
Quadro 2000     2.1
Quadro 2000D     2.1
Quadro 600     2.1
Quadro FX 5800     1.3
Quadro FX 5600     1.0
Quadro FX 4800     1.3
Quadro FX 4800 for Mac     1.3
Quadro FX 4700 X2     1.1
Quadro FX 4600     1.0
Quadro FX 3800     1.3
Quadro FX 3700     1.1
Quadro FX 1800     1.1
Quadro FX 1700     1.1
Quadro FX 580     1.1
Quadro FX 570     1.1
Quadro FX 470     1.1
Quadro FX 380     1.1
Quadro FX 380 Low Profile     1.2
Quadro FX 370     1.1
Quadro FX 370 Low Profile     1.1
Quadro CX     1.3
Quadro NVS 450     1.1
Quadro NVS 420     1.1
NVIDIA NVS 300     1.2
Quadro NVS 295     1.1
Quadro Plex 7000     2.0
Quadro Plex 2200 D2     1.3
Quadro Plex 2100 D4     1.1
Quadro Plex 2100 S4     1.0
Quadro Mobile Products GPU     Compute Capability
Quadro 5010M     2.0
Quadro 5000M     2.0
Quadro 4000M     2.1
Quadro 3000M     2.1
Quadro 2000M     2.1
Quadro 1000M     2.1
Quadro FX 3800M     1.1
Quadro FX 3700M     1.1
Quadro FX 3600M     1.1
Quadro FX 2800M     1.1
Quadro FX 2700M     1.1
Quadro FX 1800M     1.2
Quadro FX 1700M     1.1
Quadro FX 1600M     1.1
Quadro FX 880M     1.2
Quadro FX 770M     1.1
Quadro FX 570M     1.1
Quadro FX 380M     1.2
Quadro FX 370M     1.1
Quadro FX 360M     1.1
Quadro NVS 320M     1.1
Quadro NVS 160M     1.1
Quadro NVS 150M     1.1
Quadro NVS 140M     1.1
Quadro NVS 135M     1.1
Quadro NVS 130M     1.1
CUDA-Enabled NVS Products
NVS
Desktop Products GPU     Compute Capability
Quadro NVS 450     1.1
Quadro NVS 420     1.1
NVIDIA NVS 300     1.2
Quadro NVS 295     1.1
Mobile Products GPU     Compute Capability
NVS 4200M     2.1
NVS 5100M     1.2
NVS 3100M     1.2
NVS 2100M     1.2
CUDA-Enabled GeForce Products

GeForce 8, 9, 100, 200, 400-series, 500-series GPUs with a minimum of 256MB of local graphics memory.
GeForce
GeForce Desktop Products GPU     Compute Capability
GeForce GTX 560 Ti     2.1
GeForce GTX 550 Ti     2.1
GeForce GTX 460     2.1
GeForce GTS 450     2.1
GeForce GTS 450*     2.1
GeForce GTX 590     2.0
GeForce GTX 580     2.0
GeForce GTX 570     2.0
GeForce GTX 480     2.0
GeForce GTX 470     2.0
GeForce GTX 465     2.0
GeForce GTX 295     1.3
GeForce GTX 285     1.3
GeForce GTX 285 for Mac     1.3
GeForce GTX 280     1.3
GeForce GTX 275     1.3
GeForce GTX 260     1.3
GeForce GT 520     2.1
GeForce GT 440     2.1
GeForce GT 440*     2.1
GeForce GT 430     2.1
GeForce GT 430*     2.1
GeForce GT 420*     1.0
GeForce GT 240     1.2
GeForce GT 220*     1.2
GeForce 210*     1.2
GeForce GTS 250     1.1
GeForce GTS 150     1.1
GeForce GT 130*     1.1
GeForce GT 120*     1.1
GeForce G100*     1.1
GeForce 9800 GX2     1.1
GeForce 9800 GTX+     1.1
GeForce 9800 GTX     1.1
GeForce 9600 GSO     1.1
GeForce 9500 GT     1.1
GeForce 8800 GTS     1.1
GeForce 8800 GT     1.1
GeForce 8800 GS     1.1
GeForce 8600 GTS     1.1
GeForce 8600 GT     1.1
GeForce 8500 GT     1.1
GeForce 8400 GS     1.1
GeForce 9400 mGPU     1.1
GeForce 9300 mGPU     1.1
GeForce 8300 mGPU     1.1
GeForce 8200 mGPU     1.1
GeForce 8100 mGPU     1.1
GeForce 8800 Ultra     1.0
GeForce 8800 GTX     1.0
GeForce GT 340*     1.0
GeForce GT 330*     1.0
GeForce GT 320*     1.0
GeForce 315*     1.0
GeForce 310*     1.0
GeForce 9800 GT     1.0
GeForce 9600 GT     1.0
GeForce 9400GT     1.0
GeForce Notebook Products GPU     Compute Capability
GeForce GT 635M     2.1
GeForce GT 630M     2.1
GeForce 610M     2.1
GeForce GTX 580M     2.1
GeForce GTX 570M     2.1
GeForce GTX 560M     2.1
GeForce GT 555M     2.1
GeForce GT 550M     2.1
GeForce GT 540M     2.1
GeForce GT 525M     2.1
GeForce GT 520MX     2.1
GeForce GT 520M     2.1
GeForce GTX 485M     2.1
GeForce GTX 470M     2.1
GeForce GTX 460M     2.1
GeForce GT 445M     2.1
GeForce GT 435M     2.1
GeForce GT 420M     2.1
GeForce GT 415M     2.1
GeForce GTX 480M     2.0
GeForce GTS 360M     1.2
GeForce GTS 350M     1.2
GeForce GT 335M     1.2
GeForce GT 330M     1.2
GeForce GT 325M     1.2
GeForce GT 240M     1.2
GeForce G210M     1.2
GeForce 310M     1.2
GeForce 305M     1.2
GeForce GTX 285M     1.1
GeForce GTX 280M     1.1
GeForce GTX 260M     1.1
GeForce 9800M GTX     1.1
GeForce 8800M GTX     1.1
GeForce GTS 260M     1.1
GeForce GTS 250M     1.1
GeForce 9800M GT     1.1
GeForce 9600M GT     1.1
GeForce 8800M GTS     1.1
GeForce 9800M GTS     1.1
GeForce GT 230M     1.1
GeForce 9700M GT     1.1
GeForce 9650M GS     1.1
GeForce 9700M GT     1.1
GeForce 9650M GS     1.1
GeForce 9600M GT     1.1
GeForce 9600M GS     1.1
GeForce 9500M GS     1.1
GeForce 8700M GT     1.1
GeForce 8600M GT     1.1
GeForce 8600M GS     1.1
GeForce 9500M G     1.1
GeForce 9300M G     1.1
GeForce 8400M GS     1.1
GeForce G210M     1.1
GeForce G110M     1.1
GeForce 9300M GS     1.1
GeForce 9200M GS     1.1
GeForce 9100M G     1.1
GeForce 8400M GT     1.1
GeForce G105M     1.1

ORIGINAL: v1.0

UPDATE: v1.1
Re-compiled with release.

UPDATE: v1.2
Disabled debug mode to fix crashing and increase fps slightly.
Outputs min/max/avg fps.

Attached Files

Cloth_v1_2.zip (1.87 MB, 15 views)

__________________

Last edited by Fadingz; March 2nd, 2012 at 22:29..

**Hard core Rikki** · March 1st, 2012, 19:34

Error for all. Missing files?
GT440.

__________________

NGemu toolbar http://ngemu.ourtoolbar.com

Fadingz · March 1st, 2012, 19:46

Would you please try again?
I just compiled the host on windows 32.
Perhaps because I am in debug mode?
which may require external dependencies, I'll look into that.

__________________

**Hard core Rikki** · March 1st, 2012, 19:49

Same result.

__________________

NGemu toolbar http://ngemu.ourtoolbar.com

Fadingz · March 1st, 2012, 19:50

I'll try to compile a release version.
It might take some time as I have to set up release mode for all of my static libraries.

__________________

Fadingz · March 1st, 2012, 20:14

Gah, cutil, whch I ned to output te fps, is throwing me link errors when I switch platform.
I'll try to battle with it.

__________________

**@ruantec** · March 1st, 2012, 20:39

Sadly i do not have an nvidia card installed atm to test

__________________

Current development tools:

Visual C++.net, Visual C#.net
Visual VB.net, Visual Webdeveloper.net
Bloodshed Dev C++, Borland C++
Visual Basic 6

Fadingz · March 1st, 2012, 20:45

Re-compiled with release.
Please try again.

__________________

Fadingz · March 1st, 2012, 20:47

Quote:

Originally Posted by @ruantec

Sadly i do not have an nvidia card installed atm to test

I just want to collect data as it is right now, then I am gonna keyframe the simulation and offload onto game, then compare some trade offs. (keyframe approach)
Thank you regardless for dropping by :]

__________________

**Hard core Rikki** · March 1st, 2012, 20:51

Closer, not yet there

Perhaps a reminder to not target the absolute latest library versions that hardly anyone has (therefore requiring bundling).

*cudart32_41_28.dll is missing*

__________________

NGemu toolbar http://ngemu.ourtoolbar.com

**SCHUMI_4EVER** · March 2nd, 2012, 13:40

Quote:

Originally Posted by Hard core Rikki

Closer, not yet there

Perhaps a reminder to not target the absolute latest library versions that hardly anyone has (therefore requiring bundling).

*cudart32_41_28.dll is missing*

Same problem here.

__________________

Intel Core2Quad Q9550 (2.83Ghz stock) | ASUS P5Q | 2x2GB Transcend JetRam DDR2-800 | ASUS ENGTX260\HDTP\896M | Windows Vista Home Premium 64bit SP1

The Champ has retired but may his Legacy live on FOREVER !!!!

No$GBA DS FAQ 2.0

Get it right fools! The glass is HALF-EMPTY, not half-full!!!

!!! WARNING: Emulation requires a brain !!! WARNING: Emulation =/= Piracy !!!

GeominorAI · March 2nd, 2012, 13:49

Quote:

Originally Posted by SCHUMI_4EVER

Same problem here.

It is from the NVIDIA GPU Computing Toolkit.

Attached Files

cudart32_41_28.zip (141.5 KB, 11 views)

__________________
Computational intelligence must be implemented on a parallel architecture

**Hard core Rikki** · March 1st, 2012, 20:55

To be honest, it works fine with other cuda dlls renamed to that. Just FYI.

256x256 = 90-105 fps
512x512 = 30-35 fps
1024x1024 = 6-8 fps

GT440

__________________

NGemu toolbar http://ngemu.ourtoolbar.com

**SCHUMI_4EVER** · March 2nd, 2012, 13:53

Thanks Geo

Update:
256x256 = 96-123FPS (I think, was hard to tell with the rapidly changing numbers).
512x512= +/- 43FPS
1024x1024 = 6-13FPS

GTX260 as my sig says.

__________________

Intel Core2Quad Q9550 (2.83Ghz stock) | ASUS P5Q | 2x2GB Transcend JetRam DDR2-800 | ASUS ENGTX260\HDTP\896M | Windows Vista Home Premium 64bit SP1

The Champ has retired but may his Legacy live on FOREVER !!!!

No$GBA DS FAQ 2.0

Get it right fools! The glass is HALF-EMPTY, not half-full!!!

!!! WARNING: Emulation requires a brain !!! WARNING: Emulation =/= Piracy !!!

Tyo · March 2nd, 2012, 15:27

256x256 = 280-320FPS (seriously hard to notice the actual number, so i'm just randomly pause to get few fps sample. how about making a counter for min, max and average ?)
512x512= 110-114 FPS
1024x1024 = 27-31FPS

GT 560ti (pre-OC from MSI, so it's running @950MHz)

Fadingz · March 2nd, 2012, 15:45

Quote:

Originally Posted by Tyo

256x256 = 280-320FPS (seriously hard to notice the actual number, so i'm just randomly pause to get few fps sample. how about making a counter for min, max and average ?)
512x512= 110-114 FPS
1024x1024 = 27-31FPS

GT 560ti (pre-OC from MSI, so it's running @950MHz)

Good point, I'll do that.
Will also make a GUI for particle count as well.

__________________

**SCHUMI_4EVER** · March 2nd, 2012, 15:34

Oh wow...I wonder whether it's the newer CUDA version or purely your card's superior power that lead to figures so much greater than mine.

__________________

Intel Core2Quad Q9550 (2.83Ghz stock) | ASUS P5Q | 2x2GB Transcend JetRam DDR2-800 | ASUS ENGTX260\HDTP\896M | Windows Vista Home Premium 64bit SP1

The Champ has retired but may his Legacy live on FOREVER !!!!

No$GBA DS FAQ 2.0

Get it right fools! The glass is HALF-EMPTY, not half-full!!!

!!! WARNING: Emulation requires a brain !!! WARNING: Emulation =/= Piracy !!!

Tyo · March 2nd, 2012, 19:30

@schumi
i've tested both executable for cuda 1.0 and 2.0, the performance somewhat the same.
so, probably the raw power. 560ti have twice the number of cuda core than 260 (assuming your 260 is the original 192 core) and clocked higher

@fadingz
can't wait to test the next one.

btw, i'm not sure whether it's just my system or anything else.
but the app always close automatically after few seconds. and it's consistent in certain frame in every executable.
256x256 => few seconds after the cloth is gone / drop from the object
512x512 => got to se teh cloth on the object for a while until it closed
1024x1024 => couple seconds after the cloth if drop onto the object

win 7 X64, 295.73 driver

**SCHUMI_4EVER** · March 2nd, 2012, 21:43

Quote:

Originally Posted by Tyo

@schumi
i've tested both executable for cuda 1.0 and 2.0, the performance somewhat the same.
so, probably the raw power. 560ti have twice the number of cuda core than 260 (assuming your 260 is the original 192 core) and clocked higher

Yeah I have the original Core 192 version not the 216.
I was just surprised to see the performance difference considering the gap in game performance doesn't appear to be anywhere near as large.

__________________

Intel Core2Quad Q9550 (2.83Ghz stock) | ASUS P5Q | 2x2GB Transcend JetRam DDR2-800 | ASUS ENGTX260\HDTP\896M | Windows Vista Home Premium 64bit SP1

The Champ has retired but may his Legacy live on FOREVER !!!!

No$GBA DS FAQ 2.0

Get it right fools! The glass is HALF-EMPTY, not half-full!!!

!!! WARNING: Emulation requires a brain !!! WARNING: Emulation =/= Piracy !!!

Fadingz · March 2nd, 2012, 22:20

Made a new version with min/max/avg fps. v1_2
Fixed the crashing bug, forgot to set debug mode off, which should also increase the fps to a degree.

Will work on the GUI.

I ticked on L1/L2 caching for the 2.0 but it doesn't seem to matter much in this case for some reason.
I'll look into it.

__________________

Last edited by Fadingz; March 2nd, 2012 at 22:30..

Thread Tools
Show Printable Version Email this Page
Display Modes
Switch to Linear Mode Hybrid Mode Switch to Threaded Mode

March 1st, 2012, 19:34	#2
Hard core Rikki Moving into the beat Join Date: May 2004 Location: Perpetual Hawaii Posts: 11,265	Error for all. Missing files? GT440. __________________ deviantART \| Blog \| VBA-M \| GeneralEmu \| XTemu \| FMVTool2 \| @ES \| Shendo new NGemu toolbar http://ngemu.ourtoolbar.com

March 1st, 2012, 19:46	#3
Fadingz 代言人 Join Date: Dec 2006 Location: 應許之地 Posts: 7,056	Would you please try again? I just compiled the host on windows 32. Perhaps because I am in debug mode? which may require external dependencies, I'll look into that. __________________

March 1st, 2012, 19:49	#4
Hard core Rikki Moving into the beat Join Date: May 2004 Location: Perpetual Hawaii Posts: 11,265	Same result. __________________ deviantART \| Blog \| VBA-M \| GeneralEmu \| XTemu \| FMVTool2 \| @ES \| Shendo new NGemu toolbar http://ngemu.ourtoolbar.com

March 1st, 2012, 19:50	#5
Fadingz 代言人 Join Date: Dec 2006 Location: 應許之地 Posts: 7,056	I'll try to compile a release version. It might take some time as I have to set up release mode for all of my static libraries. __________________

March 1st, 2012, 20:14	#6
Fadingz 代言人 Join Date: Dec 2006 Location: 應許之地 Posts: 7,056	Gah, cutil, whch I ned to output te fps, is throwing me link errors when I switch platform. I'll try to battle with it. __________________

March 1st, 2012, 20:39	#7
@ruantec Crazy GFX coder Join Date: Nov 2002 Location: Dominican Republic/Austria Posts: 8,093	Sadly i do not have an nvidia card installed atm to test __________________ Current development tools: Visual C++.net, Visual C#.net Visual VB.net, Visual Webdeveloper.net Bloodshed Dev C++, Borland C++ Visual Basic 6

March 1st, 2012, 20:45	#8
Fadingz 代言人 Join Date: Dec 2006 Location: 應許之地 Posts: 7,056	Re-compiled with release. Please try again. __________________

March 1st, 2012, 20:51	#10
Hard core Rikki Moving into the beat Join Date: May 2004 Location: Perpetual Hawaii Posts: 11,265	Closer, not yet there Perhaps a reminder to not target the absolute latest library versions that hardly anyone has (therefore requiring bundling). cudart32_41_28.dll is missing __________________ deviantART \| Blog \| VBA-M \| GeneralEmu \| XTemu \| FMVTool2 \| @ES \| Shendo new NGemu toolbar http://ngemu.ourtoolbar.com

March 1st, 2012, 20:55	#13
Hard core Rikki Moving into the beat Join Date: May 2004 Location: Perpetual Hawaii Posts: 11,265	To be honest, it works fine with other cuda dlls renamed to that. Just FYI. 256x256 = 90-105 fps 512x512 = 30-35 fps 1024x1024 = 6-8 fps GT440 __________________ deviantART \| Blog \| VBA-M \| GeneralEmu \| XTemu \| FMVTool2 \| @ES \| Shendo new NGemu toolbar http://ngemu.ourtoolbar.com

March 2nd, 2012, 13:53	#14
SCHUMI_4EVER Hackin 'n Slashin Join Date: Jan 2007 Location: Most-expensive-Internet-on-the-Planet Land : South Africa Posts: 26,309	Thanks Geo Update: 256x256 = 96-123FPS (I think, was hard to tell with the rapidly changing numbers). 512x512= +/- 43FPS 1024x1024 = 6-13FPS GTX260 as my sig says. __________________ Intel Core2Quad Q9550 (2.83Ghz stock) \| ASUS P5Q \| 2x2GB Transcend JetRam DDR2-800 \| ASUS ENGTX260\HDTP\896M \| Windows Vista Home Premium 64bit SP1 The Champ has retired but may his Legacy live on FOREVER !!!! No$GBA DS FAQ 2.0 Get it right fools! The glass is HALF-EMPTY, not half-full!!! !!! WARNING: Emulation requires a brain !!! WARNING: Emulation =/= Piracy !!!

March 2nd, 2012, 15:27	#15
Tyo Registered User Join Date: Jan 2008 Posts: 363	256x256 = 280-320FPS (seriously hard to notice the actual number, so i'm just randomly pause to get few fps sample. how about making a counter for min, max and average ?) 512x512= 110-114 FPS 1024x1024 = 27-31FPS GT 560ti (pre-OC from MSI, so it's running @950MHz)

March 2nd, 2012, 15:34	#17
SCHUMI_4EVER Hackin 'n Slashin Join Date: Jan 2007 Location: Most-expensive-Internet-on-the-Planet Land : South Africa Posts: 26,309	Oh wow...I wonder whether it's the newer CUDA version or purely your card's superior power that lead to figures so much greater than mine. __________________ Intel Core2Quad Q9550 (2.83Ghz stock) \| ASUS P5Q \| 2x2GB Transcend JetRam DDR2-800 \| ASUS ENGTX260\HDTP\896M \| Windows Vista Home Premium 64bit SP1 The Champ has retired but may his Legacy live on FOREVER !!!! No$GBA DS FAQ 2.0 Get it right fools! The glass is HALF-EMPTY, not half-full!!! !!! WARNING: Emulation requires a brain !!! WARNING: Emulation =/= Piracy !!!

March 2nd, 2012, 19:30	#18
Tyo Registered User Join Date: Jan 2008 Posts: 363	@schumi i've tested both executable for cuda 1.0 and 2.0, the performance somewhat the same. so, probably the raw power. 560ti have twice the number of cuda core than 260 (assuming your 260 is the original 192 core) and clocked higher @fadingz can't wait to test the next one. btw, i'm not sure whether it's just my system or anything else. but the app always close automatically after few seconds. and it's consistent in certain frame in every executable. 256x256 => few seconds after the cloth is gone / drop from the object 512x512 => got to se teh cloth on the object for a while until it closed 1024x1024 => couple seconds after the cloth if drop onto the object win 7 X64, 295.73 driver

March 2nd, 2012, 22:20	#20
Fadingz 代言人 Join Date: Dec 2006 Location: 應許之地 Posts: 7,056	Made a new version with min/max/avg fps. v1_2 Fixed the crashing bug, forgot to set debug mode off, which should also increase the fps to a degree. Will work on the GUI. I ticked on L1/L2 caching for the 2.0 but it doesn't seem to matter much in this case for some reason. I'll look into it. __________________ Last edited by Fadingz; March 2nd, 2012 at 22:30..