SETI@home Version 7 has been released

Message boards : News : SETI@home Version 7 has been released
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 . . . 20 · Next

AuthorMessage
Phil

Send message
Joined: 12 Mar 04
Posts: 14
Credit: 4,841,296
RAC: 0
United Kingdom
Message 1373857 - Posted: 30 May 2013, 22:15:55 UTC

Is there a problem with v7(opencl)? I am running Cuda42 and if my wingman is also running Cuda then the results are validated. But if my wingman is running opencl then the result is marked as inconclusive and if the 2nd wingman is also running opencl then result becomes invalid but if they are running cuda then the result is valid. Any suggestions?
ID: 1373857 · Report as offensive
JB
Volunteer tester

Send message
Joined: 21 Jul 09
Posts: 46
Credit: 7,891,307
RAC: 0
Germany
Message 1373866 - Posted: 30 May 2013, 22:36:14 UTC

Look at Richards posting refered to ATIcards+opencl in this thread:

http://setiathome.berkeley.edu/forum_thread.php?id=71777

"......using the CPUs only for Multibeam ...."

ID: 1373866 · Report as offensive
Grumpy Swede
Volunteer tester
Avatar

Send message
Joined: 1 Nov 08
Posts: 7525
Credit: 46,933,516
RAC: 8,251
Sweden
Message 1373870 - Posted: 30 May 2013, 22:43:36 UTC - in response to Message 1373866.  
Last modified: 30 May 2013, 22:46:20 UTC

Look at Richards posting refered to ATIcards+opencl in this thread:

http://setiathome.berkeley.edu/forum_thread.php?id=71777

"......using the CPUs only for Multibeam ...."



Richards posts is about Boinc 6 that can't (and for now shouldn't) get ATI OpenCL MB v7 tasks at all. These SDK 2.5 clients runs Boinc 7, and they can get ATI OpenCL MB v7 tasks, but the problem is that they do get these tasks despite not having at least SDK 2.6 drivers.
WARNING!! "THIS IS A SIGNATURE", of the "IT MAY CHANGE AT ANY MOMENT" type. It may, or may not be considered insulting, all depending upon HOW SENSITIVE THE VIEWER IS, to certain inputs to/from the nervous system.
ID: 1373870 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6104
Credit: 95,724,812
RAC: 33,569
Russia
Message 1373875 - Posted: 30 May 2013, 22:55:23 UTC

it's a real problem cause false overflow occurs too fast.
Perhaps it's worth to stop plan class for older drivers at all if it can't distinguish SDK 2.5 drivers from SDK 2.6 ones.

SETI apps news
We're not gonna fight them. We're gonna transcend them.
ID: 1373875 · Report as offensive
Phil

Send message
Joined: 12 Mar 04
Posts: 14
Credit: 4,841,296
RAC: 0
United Kingdom
Message 1373876 - Posted: 30 May 2013, 22:56:42 UTC

Does this mean that apart from the waste of computer time, a lot of good results are being discarded as invalid whilst others that should be discarded are being marked as valid and therefore included in the database
ID: 1373876 · Report as offensive
Horacio

Send message
Joined: 14 Jan 00
Posts: 536
Credit: 75,967,266
RAC: 0
Argentina
Message 1373877 - Posted: 30 May 2013, 22:58:18 UTC - in response to Message 1373842.  

Both wingmen are running too low an APP runtime version, need to be an APP SDK 2.6 driver/APP runtime :

Driver version: CAL 1.4.1607 (VM)
Version: OpenCL 1.1 AMD-APP-SDK-v2.5 (793.1)

Claggy

All my inconclusives are paired with hosts using the AMD-APP-SDK-v2.5... and they all ended in -9 overflow (except one with no data in the stderr)...
It seems something that needs to be addressed ASAP...
ID: 1373877 · Report as offensive
Grumpy Swede
Volunteer tester
Avatar

Send message
Joined: 1 Nov 08
Posts: 7525
Credit: 46,933,516
RAC: 8,251
Sweden
Message 1373878 - Posted: 30 May 2013, 22:58:58 UTC - in response to Message 1373876.  
Last modified: 30 May 2013, 23:02:01 UTC

Does this mean that apart from the waste of computer time, a lot of good results are being discarded as invalid whilst others that should be discarded are being marked as valid and therefore included in the database


At the moment, that seems to be what happens yes. I do hope, for the sake of scientific accuracy, that those SDK 2.5 hosts are stopped ASAP from getting any new ATI MB v7 tasks.
WARNING!! "THIS IS A SIGNATURE", of the "IT MAY CHANGE AT ANY MOMENT" type. It may, or may not be considered insulting, all depending upon HOW SENSITIVE THE VIEWER IS, to certain inputs to/from the nervous system.
ID: 1373878 · Report as offensive
JB
Volunteer tester

Send message
Joined: 21 Jul 09
Posts: 46
Credit: 7,891,307
RAC: 0
Germany
Message 1373881 - Posted: 30 May 2013, 23:04:42 UTC - in response to Message 1373870.  

Look at Richards posting refered to ATIcards+opencl in this thread:

http://setiathome.berkeley.edu/forum_thread.php?id=71777

"......using the CPUs only for Multibeam ...."



Richards posts is about Boinc 6 that can't (and for now shouldn't) get ATI OpenCL MB v7 tasks at all. These SDK 2.5 clients runs Boinc 7, and they can get ATI OpenCL MB v7 tasks, but the problem is that they do get these tasks despite not having at least SDK 2.6 drivers.


Ah, okay.
I just looked into the STDERR report of his wingman:

" Driver version: CAL 1.4.1607 (VM)
" Version: OpenCL 1.1 AMD-APP-SDK-v2.5 (793.1)

ID: 1373881 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6104
Credit: 95,724,812
RAC: 33,569
Russia
Message 1373902 - Posted: 30 May 2013, 23:39:16 UTC

AFAIK this typo in plan class eradicated already. So look for new work assigment to those hosts. They should not get new MB v7 tasks...
SETI apps news
We're not gonna fight them. We're gonna transcend them.
ID: 1373902 · Report as offensive
Jeroen
Volunteer tester

Send message
Joined: 28 Mar 01
Posts: 13
Credit: 8,155,940
RAC: 0
United States
Message 1373942 - Posted: 31 May 2013, 2:31:01 UTC
Last modified: 31 May 2013, 2:36:16 UTC

So far my Windows 680 host and Linux 7970 host have successfully completed the new version 7 tasks. The 680 host picked up the CUDA 5 application. I just built an app_info.xml file for my Linux host so that I can run both setiathome_7.04_x86_64-pc-linux-gnu__opencl_ati5_linux and ap_6.07r1844_sse2_clATI_linux64. GPU load is in the range of 91-93% on both cards.
ID: 1373942 · Report as offensive
Profile Pendleton

Send message
Joined: 29 Dec 01
Posts: 5
Credit: 8,684,741
RAC: 0
United States
Message 1374012 - Posted: 31 May 2013, 6:02:12 UTC

Thanks!!!! :) ;)
ID: 1374012 · Report as offensive
triplemmm

Send message
Joined: 21 Nov 00
Posts: 27
Credit: 7,152,794
RAC: 5,364
Australia
Message 1374074 - Posted: 31 May 2013, 7:37:22 UTC

i dont know what is wrong but i have the new version 7 and now 118 errors in download from the 3021448212 1256148716 5879391 30 May 2013, 9:22:09 UTC 30 May 2013, 10:42:06 UTC Error while downloading 0.00 0.00 --- SETI@home v7 v7.03 (opencl_ati5_sah)
hmmm need to sort this out lucky im still crunching the old system
ID: 1374074 · Report as offensive
kittyman Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 50374
Credit: 980,815,497
RAC: 50,098
United States
Message 1374079 - Posted: 31 May 2013, 7:43:23 UTC - in response to Message 1374074.  

i dont know what is wrong but i have the new version 7 and now 118 errors in download from the 3021448212 1256148716 5879391 30 May 2013, 9:22:09 UTC 30 May 2013, 10:42:06 UTC Error while downloading 0.00 0.00 --- SETI@home v7 v7.03 (opencl_ati5_sah)
hmmm need to sort this out lucky im still crunching the old system

This is a news thread......
If you would like help with your difficulty, please post your question in the Number Crunching thread.
Lots of folks there working on various questions with the v7 rollout.

Meow!
"Learn from yesterday. Live for today. Hope for tomorrow." Albert Einstein
"With cats." kittyman

ID: 1374079 · Report as offensive
Thomas
Volunteer tester

Send message
Joined: 9 Dec 11
Posts: 1499
Credit: 1,345,576
RAC: 0
France
Message 1374084 - Posted: 31 May 2013, 7:50:24 UTC - in response to Message 1374080.  

I have version 7 also, and its still crunching but there is
no New Work, and normally when I upload right away I'll get some
new work. This time my guf is almost empty. :(

Same here
ID: 1374084 · Report as offensive
kittyman Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 50374
Credit: 980,815,497
RAC: 50,098
United States
Message 1374085 - Posted: 31 May 2013, 7:50:47 UTC - in response to Message 1374080.  

I have version 7 also, and its still crunching but there is
no New Work, and normally when I upload right away I'll get some
new work. This time my guf is almost empty. :(

Again, for all here.
This is a news thread. Please post questions or problems in Number Crunching.

The kitties have their v7 caches full, so if you are properly set up, I think you should be getting enough work.

Get thee hence to the proper thread, and somebody shall try to help you figure out what is going wrong in your case.
"Learn from yesterday. Live for today. Hope for tomorrow." Albert Einstein
"With cats." kittyman

ID: 1374085 · Report as offensive
Profile Sutaru Tsureku
Volunteer tester

Send message
Joined: 6 Apr 07
Posts: 7104
Credit: 147,313,424
RAC: 0
Germany
Message 1374223 - Posted: 31 May 2013, 12:52:18 UTC

It's well known that GTX560Ti's had probs with stock SAH Enhanced apps.

But also with the new SAHv7 (x41zc) apps?

Only errors: hostid=6598140.

What is here the problem?


* Best regards! :-) * Philip J. Fry, team seti.international founder. * Optimize your PC for higher RAC. * SETI@home needs your help. *
ID: 1374223 · Report as offensive
Profile Sutaru Tsureku
Volunteer tester

Send message
Joined: 6 Apr 07
Posts: 7104
Credit: 147,313,424
RAC: 0
Germany
Message 1374234 - Posted: 31 May 2013, 13:07:29 UTC - in response to Message 1374220.  
Last modified: 31 May 2013, 13:08:42 UTC

What a nice message.

You read already the number crunching forum because of NVidia driver problems which cause computation errors (Kepler cards unsupported, the solution/workround)?


* Best regards! :-) * Philip J. Fry, team seti.international founder. * Optimize your PC for higher RAC. * SETI@home needs your help. *
ID: 1374234 · Report as offensive
Profile Gone with the wind Crowdfunding Project Donor*Special Project $75 donor
Volunteer tester

Send message
Joined: 19 Nov 00
Posts: 41574
Credit: 41,951,956
RAC: 12
Message 1374278 - Posted: 31 May 2013, 14:08:40 UTC

He has only made 1 post since 1999, I wouldn't take it too seriously!

ID: 1374278 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 12987
Credit: 138,201,271
RAC: 56,671
United Kingdom
Message 1374289 - Posted: 31 May 2013, 14:18:00 UTC - in response to Message 1374278.  

He has only made 1 post since 1999, I wouldn't take it too seriously!

Just noting in passing that this is the News thread about the release of a SETI application that will run on his Nvidia 600 card.....
ID: 1374289 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1374295 - Posted: 31 May 2013, 14:21:46 UTC - in response to Message 1374223.  
Last modified: 31 May 2013, 14:23:14 UTC

What is here the problem?


'Midrange enthusiast card' = price/performance market = 'rush to market' = 'got the voltage wrong & clocks too high---> Greedy manufacturers'

[Throw in some overeager overclockers, and the odd bad driver & system, and there you have it, 'perfect storm']

Having said that, my eVGA SC 560ti has performed flawlessly since out of the box.
6.10 is indeed very outdated & had more than its fair share of issues, but even the much improved x41zc won't fix a broken computer.

As for the guy selling his hardware off cheap, let me know how much ;)
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 1374295 · Report as offensive
Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 . . . 20 · Next

Message boards : News : SETI@home Version 7 has been released


 
©2019 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.