Eric J KorpelaVolunteer moderator
Project administrator
Project developer
Project scientist
Send message
Joined: 15 Mar 05
Posts: 1007
Credit: 6,404,320
RAC: 11,571
Few changes to science. This is mainly in preparation for v7 GPU testing. 6.98 (non-GPU) for MacOS should be appearing shortly.
____________
Great news, Eric! Looking forward to the Win x64 release!
____________
Per aspera, ad astra!
Ok is OSX GPU also coming? I hope so. Will be watching closely.
[seti.international] Philip J. FryVolunteer tester
Send message
Joined: 7 Jun 09
Posts: 234
Credit: 1,603,894
RAC: 13
Great news, Eric! Looking forward to the Win x64 release!
I don't think SETI@home and -Beta will ever release a stock 64bit Windows application.
The current apps at S@h Beta.
The current apps at S@h.
Only 32bit Windows apps.
Linux have 32 and 64bit apps.
If you would like to have 64bit Windows apps you must use the Lunatics Installer (3rd party apps).
- Best regards! :-) - Sutaru Tsureku, team seti.international founder. - Optimize your PC for higher RAC (@ SETI@home Main). - SETI@home needs your help. -
____________
>Das Deutsche Cafe. The German Cafe.<
Funny, as I am only processing with a GPU on a PC with Win 7 Ultimate x64... But hey, I'm no rocket scientist after all...
Cheers!
____________
Per aspera, ad astra!
It is not necessary for the projects to have a 64bit version to utilise 64bit systems.
The BM will just use the 32bit version if no 64bit is available.
____________
Christoph
As long as I can keep burning my GPU, that's fine with me. I was just hoping that since all apps are still using (some) CPU, I could make good use of its 64-bit capability.
Cheers!
____________
Per aspera, ad astra!
[seti.international] Philip J. FryVolunteer tester
Send message
Joined: 7 Jun 09
Posts: 234
Credit: 1,603,894
RAC: 13
As long as I can keep burning my GPU, that's fine with me. I was just hoping that since all apps are still using (some) CPU, I could make good use of its 64-bit capability.
Cheers!
The stock 32bit applications work well on a 64bit operating system.
You know the Lunatics Installer which will install optimized project applications for SETI@home?
The optimized project applications which use the CPU instructions (SSE*/**, SSE2*/**, SSE3*, SSSE3*, SSE4.1* or AVX**) run ~ 50 % faster than the stock applications which BOINC download automatically from the project.
[* SETI@home Enhanced]
[** AstroPulse]
As far as I know, the speed difference between 32 and 64bit optimized applications is only ~ 5 % at SETI@home (more is not possible).
Also, you could calculate then on your AMD/ATI GPU AstroPulse and SETI@home Enhanced work units.
For more infos we could talk via private message or in the NC subforum at SETI@home. :-)
- Best regards! :-) - Sutaru Tsureku, team seti.international founder. - Optimize your PC for higher RAC (@ SETI@home Main). - SETI@home needs your help. -
____________
>Das Deutsche Cafe. The German Cafe.<
I have an est runtime on my tasks of 18hours or something.
And they take like 6 hours to complete.
http://setiweb.ssl.berkeley.edu/beta/results.php?hostid=55958&offset=0&show_names=0&state=2&appid=
It is just me here again.
Nothing of big importance.
Is this common??
I've got a few of those on my E8500/GTX9800+ host, they are just lost Wu's getting to the head of the cue again, see this (locked) thread:
WU download error: couldn't get input files
I did try and download them manually, but the files are no longer where they were supposed to be:
http://boinc2.ssl.berkeley.edu/beta/download/247/05oc11al.5666.18100.6.14.106
http://boinc2.ssl.berkeley.edu/beta/download/12e/05oc11al.5666.18100.6.14.101
http://boinc2.ssl.berkeley.edu/beta/download/247/05oc11al.5666.18100.6.14.100
Claggy
Maybe they are send with UDP instead of TCP.
That was intended to be a joke. I don't think they can be send with UDP.
Someone should take a look at where they are going or if they are there in the first place. Maybe they are just "tags"(ln -s).
My bad english, I haven't a clue of what it is called so I call them tags.
Tag could be the file name and not the actual file....or something like that.
[seti.international] Philip J. FryVolunteer tester
Send message
Joined: 7 Jun 09
Posts: 234
Credit: 1,603,894
RAC: 13
My BOINC couldn't download S@h v7 WUs ..
15 download errors ..
I set *no new tasks* in BOINC .. until someone report that the download go again well .. I don't want to destroy further the daily quota of new WUs ..
- Best regards! :-) - Sutaru Tsureku, team seti.international founder. - Optimize your PC for higher RAC (@ SETI@home Main). - SETI@home needs your help. -
____________
>Das Deutsche Cafe. The German Cafe.<
Just reported 53 download errored tasks, all resends, now down to a quota of 1 a day,
Claggy
Hi everyone!
I finally got a bit more time on my hands as well as constant online access, so I'm back:) If anybody needs my Atom/ION combo for testing anything please let me know. I kinda gave up on doing OpenCL on the ION but if someone wants me to (Raistmer?) I'd be more than happy to.
Anyway, I've had a few CPU tasks for days now but they refuse to start crunching by themselves. So I just now suspended my tasks from main so I could get these started. Can anybody guess why this might be happening, or does someone know a better way of getting around this so I don't have to suspend/resume tasks ie babysit Boinc?
Just reported 53 download errored tasks, all resends, now down to a quota of 1 a day,
Claggy
Likewise (though not so many). In every case I've checked, it is indeed the individual data file which is non-downloadable - I'm away from home, and unable to check exactly what reason is given in the event log.
But they are all very old tasks - in some cases, first sent out for crunching last year (December 2011). I'm assuming the files were deleted from WU storage during one of the clear-outs since then.
Edit - managed to get remote access to the log:
24/09/2012 09:12:51 | SETI@home Beta Test | [sched_op] Reason: Unrecoverable error for task 05oc11al.25211.476.7.14.130_2 (WU download error: couldn't get input files:<file_xfer_error> <file_name>05oc11al.25211.476.7.14.130</file_name> <error_code>-224</error_code> <error_message>permanent HTTP error</error_message></file_xfer_error>)
Now i've managed to get my E8500/9800GTX+ host (as well as the i7-2600K) down to 1Wu/day with 43 download errors, all resends, only two wu's downloaded O.K,
It has enough Wu's on hand to drive the limit back up over the next couple of days.
Claggy
Boy, I think I may have bitten more than I can chew here...
Seriously, I'm feeling way in over my head. So I will be happy just running my WUs and help with the project. I don't want to go installing third party code on any of my machines. Stock will have to do. After all, I think that this is what we are trying to achieve here: running whatever will the end user run at the main project. That's also why I don't think that altering the .xml files or tampering with other files from the WU is doing any good.
Thanks anyway! Cheers!
____________
Per aspera, ad astra!
[seti.international] Philip J. FryVolunteer tester
Send message
Joined: 7 Jun 09
Posts: 234
Credit: 1,603,894
RAC: 13
Hi everyone!
I finally got a bit more time on my hands as well as constant online access, so I'm back:) If anybody needs my Atom/ION combo for testing anything please let me know. I kinda gave up on doing OpenCL on the ION but if someone wants me to (Raistmer?) I'd be more than happy to.
Anyway, I've had a few CPU tasks for days now but they refuse to start crunching by themselves. So I just now suspended my tasks from main so I could get these started. Can anybody guess why this might be happening, or does someone know a better way of getting around this so I don't have to suspend/resume tasks ie babysit Boinc?
I think the primary test is SETI@home v7 now.
Since the CPU and GPU AstroPulse applications was released at SETI@home.
This could be because of a few reasons ..
Maybe the deadline of the other work units is shorter.
The project share of SETI@home Beta is too low in your BOINC.
And so on ..
Boy, I think I may have bitten more than I can chew here...
Seriously, I'm feeling way in over my head. So I will be happy just running my WUs and help with the project. I don't want to go installing third party code on any of my machines. Stock will have to do. After all, I think that this is what we are trying to achieve here: running whatever will the end user run at the main project. That's also why I don't think that altering the .xml files or tampering with other files from the WU is doing any good.
Thanks anyway! Cheers!
Here at SETI@home Beta we test the stock SETI@home, but also the 3rd party applications of the Lunatics crew.
After successful test here at SETI@home Beta and release at SETI@home, BOINC will download the stock applications automatically then at SETI@home.
If you would like that the science will be calculated faster, you would need to install the Lunatics optimized applications manually with help of an installer software, which is very easy.
[EDIT: lunatics.kwsn.net]
- Best regards! :-) - Sutaru Tsureku, team seti.international founder. - Optimize your PC for higher RAC (@ SETI@home Main). - SETI@home needs your help. -
____________
>Das Deutsche Cafe. The German Cafe.<
Alex Storey wrote:
[snip]...I've had a few CPU tasks for days now but they refuse to start crunching by themselves. So I just now suspended my tasks from main so I could get these started. Can anybody guess why this might be happening, or does someone know a better way of getting around this so I don't have to suspend/resume tasks ie babysit Boinc?
This could be because of a few reasons ..
Maybe the deadline of the other work units is shorter.
The project share of SETI@home Beta is too low in your BOINC.
And so on ..
Thanx for the reply Sutaru. Only other project I crunch is Seti and it isn't running in high priority. Also the resource share is unchanged since the beginning of the Summer when I stopped doing Beta, so I don't think that's the problem either. However I set Beta to No New Tasks all Summer long... Should I have detached and reattached?