ATI Catalyst Driver 11.12

Message boards : Number crunching : ATI Catalyst Driver 11.12
Message board moderation

To post messages, you must log in.

AuthorMessage
Profile TRuEQ & TuVaLu
Volunteer tester
Avatar

Send message
Joined: 4 Oct 99
Posts: 505
Credit: 69,523,653
RAC: 10
Sweden
Message 1178940 - Posted: 17 Dec 2011, 20:35:06 UTC

I am running ATI Catalyst Driver 11.12 now and it works for me on both MB and ap tasks using lunatics unified installer 0.39

Running 2ap or 2mb simultaniosly

Link to Computor: http://setiathome.berkeley.edu/results.php?hostid=6265988
TRuEQ & TuVaLu
ID: 1178940 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34255
Credit: 79,922,639
RAC: 80
Germany
Message 1178969 - Posted: 17 Dec 2011, 22:13:59 UTC

Yes thats already confirmed.



With each crime and every kindness we birth our future.
ID: 1178969 · Report as offensive
Profile TRuEQ & TuVaLu
Volunteer tester
Avatar

Send message
Joined: 4 Oct 99
Posts: 505
Credit: 69,523,653
RAC: 10
Sweden
Message 1179034 - Posted: 18 Dec 2011, 8:16:53 UTC - in response to Message 1178969.  

Yes thats already confirmed.


I must have missed that thread......
TRuEQ & TuVaLu
ID: 1179034 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1179049 - Posted: 18 Dec 2011, 10:54:37 UTC - in response to Message 1179034.  

Yes thats already confirmed.


I must have missed that thread......

The thread is only viewerable by Lunatics Devs and Alpha Testers, so you wouldn't have seen it.

Claggy
ID: 1179049 · Report as offensive
Profile TRuEQ & TuVaLu
Volunteer tester
Avatar

Send message
Joined: 4 Oct 99
Posts: 505
Credit: 69,523,653
RAC: 10
Sweden
Message 1179052 - Posted: 18 Dec 2011, 11:43:13 UTC - in response to Message 1179049.  

Yes thats already confirmed.


I must have missed that thread......

The thread is only viewerable by Lunatics Devs and Alpha Testers, so you wouldn't have seen it.

Claggy



That explains it. :)

Then my thread in this Forum wasn't such a bad idea after all....I guess

Any speed tests done in the other forum that can be shared?
TRuEQ & TuVaLu
ID: 1179052 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1179057 - Posted: 18 Dec 2011, 12:21:53 UTC - in response to Message 1179052.  

Any speed tests done in the other forum that can be shared?


Here are the bench times on my i7-2600K/GTX460/HD5770 host:

Cat 11.5/OpenCL 1.1 AMD-APP-SDK-v2.4 (595.10):

App Name					Task name	AR		CPU time	Elapsed
 MB7_win_x86_SSE3_OpenCL_ATi_HD5_r390		 PG0009.wu 	0.008955	12.589		161.835	
 MB7_win_x86_SSE3_OpenCL_ATi_LHD4K_r390		 PG0009.wu 	0.008955	14.539		165.891	
 MB7_win_x86_SSE3_OpenCL_ATi_r390		 PG0009.wu 	0.008955	12.386		160.711	

 MB7_win_x86_SSE3_OpenCL_ATi_HD5_r390		 PG0395.wu 	0.394768	11.653		101.587	
 MB7_win_x86_SSE3_OpenCL_ATi_LHD4K_r390		 PG0395.wu 	0.394768	13.588		146.983	
 MB7_win_x86_SSE3_OpenCL_ATi_r390		 PG0395.wu 	0.394768	11.528		140.806	

 MB7_win_x86_SSE3_OpenCL_ATi_HD5_r390		 PG0444.wu 	0.444184	11.809		88.327	
 MB7_win_x86_SSE3_OpenCL_ATi_LHD4K_r390		 PG0444.wu 	0.444184	13.931		130.915	
 MB7_win_x86_SSE3_OpenCL_ATi_r390		 PG0444.wu 	0.444184	11.591		128.435	

 MB7_win_x86_SSE3_OpenCL_ATi_HD5_r390		 PG1327.wu 	1.326684	9.126		47.050	
 MB7_win_x86_SSE3_OpenCL_ATi_LHD4K_r390		 PG1327.wu 	1.326684	11.045		70.263	
 MB7_win_x86_SSE3_OpenCL_ATi_r390		 PG1327.wu 	1.326684	9.220		64.350	


Cat 11.12/OpenCL 1.1 AMD-APP (831.4)

App Name					Task name	AR		CPU time	Elapsed
 MB7_win_x86_SSE3_OpenCL_ATi_HD5_r390		 PG0009.wu 	0.008955	11.669		142.038	
 MB7_win_x86_SSE3_OpenCL_ATi_LHD4K_r390		 PG0009.wu 	0.008955	13.057		135.549	
 MB7_win_x86_SSE3_OpenCL_ATi_r390		 PG0009.wu 	0.008955	11.107		131.118
	
 MB7_win_x86_SSE3_OpenCL_ATi_HD5_r390		 PG0395.wu 	0.394768	10.358		84.334	
 MB7_win_x86_SSE3_OpenCL_ATi_LHD4K_r390		 PG0395.wu 	0.394768	12.090		106.408	
 MB7_win_x86_SSE3_OpenCL_ATi_r390		 PG0395.wu 	0.394768	10.171		103.791
	
 MB7_win_x86_SSE3_OpenCL_ATi_HD5_r390		 PG0444.wu 	0.444184	10.733		75.910	
 MB7_win_x86_SSE3_OpenCL_ATi_LHD4K_r390		 PG0444.wu 	0.444184	12.059		97.594	
 MB7_win_x86_SSE3_OpenCL_ATi_r390		 PG0444.wu 	0.444184	10.546		94.770
	
 MB7_win_x86_SSE3_OpenCL_ATi_HD5_r390		 PG1327.wu 	1.326684	8.206		43.337	
 MB7_win_x86_SSE3_OpenCL_ATi_LHD4K_r390		 PG1327.wu 	1.326684	9.968		64.851	
 MB7_win_x86_SSE3_OpenCL_ATi_r390		 PG1327.wu 	1.326684	8.362		57.268	


Claggy
ID: 1179057 · Report as offensive
Profile TRuEQ & TuVaLu
Volunteer tester
Avatar

Send message
Joined: 4 Oct 99
Posts: 505
Credit: 69,523,653
RAC: 10
Sweden
Message 1179066 - Posted: 18 Dec 2011, 14:21:44 UTC

Thank you Claggy.


TRuEQ & TuVaLu
ID: 1179066 · Report as offensive
Profile skildude
Avatar

Send message
Joined: 4 Oct 00
Posts: 9541
Credit: 50,759,529
RAC: 60
Yemen
Message 1179144 - Posted: 18 Dec 2011, 21:48:05 UTC - in response to Message 1179066.  

Could you test the 11.2 drivers against the 11.12 to see if they are an actual improvement


In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope
ID: 1179144 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1179169 - Posted: 19 Dec 2011, 0:22:39 UTC - in response to Message 1179144.  

Could you test the 11.2 drivers against the 11.12 to see if they are an actual improvement

I know going from Cat 11.2 to Cat 11.3 took 20 minutes off ATI r521 AP tasks on my HD5770, down from 1 hr 50mins to 1 hr 30 mins,
I haven't been all the way down to Cat 11.2 in many many months, so i have no benches of ATI_r390 with Cat 11.2,
but since with different ATI GPUs 'Your Mileage May Vary', it might be better that you run benches on your HD58** and HD69** ATI's,
the latest 'Knabench 2.08 for Multibeam without Applications' is available here:

Test and Benchmark Tools

you'll also want the PG shortened test Wu's from the 'MB test WU s 2/2' package, the Stock 6.03 app and dll's (as the reference app) and the app(s) you're going to test, further instructions are in the readme,

Claggy
ID: 1179169 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34255
Credit: 79,922,639
RAC: 80
Germany
Message 1179208 - Posted: 19 Dec 2011, 8:31:32 UTC

11.5 was faster than 11.2 on my 5850 for example.
So i imagine 11.12 should be faster than 11.2. as well.
But like Claggy said it might vary on some configurations.



With each crime and every kindness we birth our future.
ID: 1179208 · Report as offensive
Profile TRuEQ & TuVaLu
Volunteer tester
Avatar

Send message
Joined: 4 Oct 99
Posts: 505
Credit: 69,523,653
RAC: 10
Sweden
Message 1179287 - Posted: 19 Dec 2011, 16:06:02 UTC
Last modified: 19 Dec 2011, 16:07:09 UTC

Thanks for the Knabench link Claggy.

I installed 12.1 preview and tested 2 WU with Knabench against 11.12
No difference at all. I guess it is the same APP (OpenCl) in them.

11.12 feels smoother in windows i think....
TRuEQ & TuVaLu
ID: 1179287 · Report as offensive
Profile Lookers

Send message
Joined: 22 Jun 99
Posts: 105
Credit: 14,840,260
RAC: 0
United Kingdom
Message 1181173 - Posted: 28 Dec 2011, 20:30:04 UTC

OK, so can somebody please give me a straightforward way of getting an Asus ATI Radeon HD6770 DirectCU running SETI ? On first installation I managed to get SETI to crunch WU's at about 5 minutes per WU. Somewhere along the line it would seem I screwed up and only succeeded in geting error messages. Have done a system restore to try and resolve issue with no success.

I am downloading WU's, but none for the Radeon GPU.

Have latest ATI driver driver (11.12) and have installed "ATI OpenCL MultiBeam 6.10 App (rev177) for Windows" from Lunatics.

Running Windows XP Pro 64 bit with an AMD Phenom 6 core 1100T and 16Gb RAM.

I cannot understand the lack of support for ATI/AMD on SETI. Throw in an Nvidia as I did (albeit on a lowly 220) and off it goes without a hitch - put in an ATI/AMD and all sorts of issues raise their ugly heads. Why is there such a lack of support for these cards on SETI - probably half the bloody world uses them.
Do your bit for science (Cancer, ET, Climate Prediction) - come and help us out.
ID: 1181173 · Report as offensive
Profile skildude
Avatar

Send message
Joined: 4 Oct 00
Posts: 9541
Credit: 50,759,529
RAC: 60
Yemen
Message 1181175 - Posted: 28 Dec 2011, 20:39:53 UTC - in response to Message 1181173.  

You appear to have a lot of ATI GPU work onboard for yourself. I've had a hard time getting work for my GPU as well. It's most likely a scheduling problem.


In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope
ID: 1181175 · Report as offensive
Profile Lookers

Send message
Joined: 22 Jun 99
Posts: 105
Credit: 14,840,260
RAC: 0
United Kingdom
Message 1181182 - Posted: 28 Dec 2011, 20:46:01 UTC

It's most likely a scheduling problem.


Thanks, I'll wait and see what happens.

Nonetheless, it would be nice to know why there is such a woeful lack of support on SETI for ATI/AMD cards. Whilst people may not throw money at the project, they put their hands in their pockets and spend their hard earned cash to buy these cards to help find SETI - so why such apathy from Berkeley ?
Do your bit for science (Cancer, ET, Climate Prediction) - come and help us out.
ID: 1181182 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1181184 - Posted: 28 Dec 2011, 20:47:47 UTC - in response to Message 1181182.  
Last modified: 28 Dec 2011, 20:48:44 UTC

It's most likely a scheduling problem.


Thanks, I'll wait and see what happens.

Nonetheless, it would be nice to know why there is such a woeful lack of support on SETI for ATI/AMD cards. Whilst people may not throw money at the project, they put their hands in their pockets and spend their hard earned cash to buy these cards to help find SETI - so why such apathy from Berkeley ?

AMD/ATI didn't write an app for Setiathome like Nvidia did, so no Stock ATI Setiathome app,

Claggy
ID: 1181184 · Report as offensive
Profile Lookers

Send message
Joined: 22 Jun 99
Posts: 105
Credit: 14,840,260
RAC: 0
United Kingdom
Message 1181186 - Posted: 28 Dec 2011, 20:52:18 UTC
Last modified: 28 Dec 2011, 20:53:59 UTC

AMD/ATI didn't write an app for Setiathome like Nvidia did


Ah, probably explains a lot. I'll persevere with SETI having been doing it since 1999 and see if I can resolve the problem. Other projects are worthwhile - but for my sins I am a SETI addict :-)

Thanks
Do your bit for science (Cancer, ET, Climate Prediction) - come and help us out.
ID: 1181186 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1181188 - Posted: 28 Dec 2011, 20:56:51 UTC - in response to Message 1181186.  
Last modified: 28 Dec 2011, 21:00:58 UTC

AMD/ATI didn't write an app for Setiathome like Nvidia did


Ah, probably explains a lot. I'll persevere with SETI having been doing it since 1999 and see if I can resolve the problem. Other projects are worthwhile - but for my sins I am a SETI addict :-)

Thanks

One of your hosts has reverted to using the Stock CPU app, you'll have to reinstall the apps to get your HD5770 crunching again, and i haven't seen any completed ATI GPU tasks on that host eithier valid or pending, so you probably need to reinstall Cat 11.12 as well.

Claggy
ID: 1181188 · Report as offensive
Profile skildude
Avatar

Send message
Joined: 4 Oct 00
Posts: 9541
Credit: 50,759,529
RAC: 60
Yemen
Message 1181189 - Posted: 28 Dec 2011, 21:01:28 UTC - in response to Message 1181188.  

I think 2 of the computers listed are the same computer


In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope
ID: 1181189 · Report as offensive

Message boards : Number crunching : ATI Catalyst Driver 11.12


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.