Jump to content
mitch_de

OpenCL Oceanwave Bench and (new) CompuBench CL

374 posts in this topic

Recommended Posts

Asus GTX580 Direct CU II 10.7.2 (11C62)

 

run:

ESC key abort / [GPU] Compute: 2.40 ms Display: 417.48 fps

 

pause:

Starting event loop...

ESC key abort / [GPU] Compute: 1.41 ms Display: 708.04 fps

 

Oh, the first one was with Multisampling, this is without, was hard to catch the text because of heavy flickering, see title for compute time and fps.

 

pausez.png

runyp.png

Share this post


Link to post
Share on other sites
Advertisement

10.7.2 11C62

 

EVGA GTX 460 SuperClocked with patched GeForceGLDriver & libclh.dylib

 

No multisampling:

 

173,18 fps - 5,77 ms

290,91 fps - 3,44 ms (paused)

Share this post


Link to post
Share on other sites

Gigabyte HD5970 ( single GPU enabled )

 

153 FPS / 698 FPS

 

whats wrong here? I got 155/810 FPS with my Sapphire Toxic HD4890, while with more powerful HD5970 ( with single GPU enabled , its still powerful than 4890 ) I get 153/698 FPS! :D

Share this post


Link to post
Share on other sites

It can be that the OpenCL compiler (complies on the fly at runtime) better knows 47xx gnu than 5xxx gnu type and so better produces the gnu program.

Also it seems that Nvidia gpus GTX get much faster gnu code than AMDs. Even the high end AMDs are faster "on the paper".

Last but not least it can be that both gpus run at different PCI speed (count of lanes for data transfer). Alls complex CUDA + OpenCL apps are bond by transfer speed CPU GPU beside VRAM speed and gnu speed - and that of core isn't slow with AMD 4890 compared to 5xxx gnu.

Share this post


Link to post
Share on other sites
Screen_Shot_2011_09_28_at_12.28.33_PM.png

10.7.1

.... Error opening file ....

Please so the steps in the how to .

STart Terminal cd (change directory) to the folder with the files content, start the app by drag & drop into the terminal + ENTER

Simple double click doenst work and gives that error you posted.

Share this post


Link to post
Share on other sites

10.7.2 11C71

 

EVGA GTX 460 SuperClocked with patched GeForceGLDriver & libclh.dylib

 

No multisampling:

 

172,12 fps - 5,84 ms

289,09 fps - 3,53 ms (paused)

Share this post


Link to post
Share on other sites

Asus Laptop G74SX 16GB memory GTX 560M 3GB

 

Starting event loop...

ESC key abort / [GPU] Compute: 6.43 ms Display: 155.60 fps

 

Not very look good ha?

 

 

Lion 10.7.1

Share this post


Link to post
Share on other sites
hi there, the open cl patches doesn't work anymore on 10.7.2 / 11C73GM

any hints or tricks?!?!?thanx folk!!!!

i tried this in 10.7.2 11C73 GM, and no problemo at all

maybe theres something wrong with your hardware

Share this post


Link to post
Share on other sites

same setup as before....rechecked hex edits...all okay!!!!

dont know!??!?!

 

!!!solved!!!

 

i forgot to edit AppleGraphicsPowerManagement.kext with correct device id!!!

 

ALL GOOD NOW!!!!thanx anyway ;-)

Share this post


Link to post
Share on other sites
same setup as before....rechecked hex edits...all okay!!!!

dont know!??!?!

 

!!!solved!!!

 

i forgot to edit AppleGraphicsPowerManagement.kext with correct device id!!!

 

ALL GOOD NOW!!!!thanx anyway ;-)

.

 

 

I did not know you could actually insert your Graphics ID into the Info.Plist. But what would be the values to enter <strings>? Did you just copy a set that was close to the same Device as yours and then just placed that section into the Info.plist?

 

Does this help prevent Kernel Panic or freezing of the screen? I just renamed it so I am not using it but if its better to have the device ID in there with the proper strings I rather do that than disabling that kext. I have to look for a thread on this.

Share this post


Link to post
Share on other sites

Upgraded to 10.7.2 GM (11C73).

 

AMD Radeon HD 6870 (MSI brand, 1GB VRAM)

 

Multisampling:

 

post-413183-1318205676_thumb.png

 

No multisampling:

 

post-413183-1318205733_thumb.png

 

Paused. no multisampling:

 

post-413183-1318205959_thumb.png

Share this post


Link to post
Share on other sites

I got this errors :(

 

Connecting to NVIDIA GeForce GTX 550 Ti...

Error opening file compute_kernels.cl

Segmentation fault: 11

Share this post


Link to post
Share on other sites
I got this errors :D

Connecting to NVIDIA GeForce GTX 550 Ti...

Error opening file compute_kernels.cl

Segmentation fault: 11

Dont start the command line app by double click ! You must run it from the terminal & set directory to folder of the app before by cd

Share this post


Link to post
Share on other sites

Can You tell me how to?

 

I tried with Terminal but he shows me the same error =(

federico-pc:~ federicocicognini$ cd /

federico-pc:/ federicocicognini$ ls

federico-pc:/ federicocicognini$ cd OpenCL_OceanWave

federico-pc:OpenCL_OceanWave federicocicognini$ open OpenCL_OceanWave

Share this post


Link to post
Share on other sites
Can You tell me how to?

 

I tried with Terminal but he shows me the same error =(

federico-pc:~ federicocicognini$ cd /

federico-pc:/ federicocicognini$ ls

federico-pc:/ federicocicognini$ cd OpenCL_OceanWave

federico-pc:OpenCL_OceanWave federicocicognini$ open OpenCL_OceanWave

Here, unzip this into OpenCL folder and double click it.

 

Share this post


Link to post
Share on other sites

Ok thanks for the help you gave to me... but the luck is not with me =( another error:

 

/Users/federicocicognini/Downloads/OceanWave_NO_Multisampling/RunMe.command: line 4: 295 Illegal instruction: 4 ./OpenCL_OceanWave

logout

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.

×