LINUX Beta Test App 4.55 available
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4312
Credit: 250396267
RAC: 34834
9 Apr 2006 11:04:42 UTC
Topic 191059
(moderation:
As some of the results of the previous 4.51 Beta Test App failed validation, I put this one with slightly modified vector code (non-SSE machines should not be affected) on the Beta Test Page. Please test and report!
BM
BM
LINUX Beta Test App 4.55 available
)
Hi Bernd !
I had crunched a few WUs with albert_451 when I saw this post. Today, I downloaded and installed albert_455 on my 2 boxes 55316 and 553623.
I have now quiesced E@H (no new work), pending validation of these results. By the time the work queues are processed, there will be 13 results ready for (quorum) validation.
Regards.
-rg-
Edited :
Reported CPU usage 20'450 -> 11'250 , or a 1.8 performance improvement !!!
Wonderful, Akosf has triggered a great new trend.
... but this also means 1.8 times smaller credits. I started a thread on the BOINC core client forum, maybe they can also produce miracles there ...
Waiting for validation of
)
Waiting for validation of first result from my computer running albert v4.55
Dual P3 Coppermine 700 @ 933
Initial estimate: 11,811 seconds
Actual time: 6449.03 seconds
I'd call this a major performance improvement.
Nothing travels faster than the speed of light with the possible exception of bad news, which obeys its own special laws.
Douglas Adams (1952 - 2001)
Linux 2.4.20-8smp, Xeon
)
Linux 2.4.20-8smp, Xeon 2.4G
http://einsteinathome.org/host/557396
failed
http://einsteinathome.org/task/24687393
http://einsteinathome.org/task/24687328
I got the first 7 results
)
I got the first 7 results witch are all valid(outcome=success).
Five of them are pending and two got credits, so the precision of SSE calculations seems to be sufficient now.
The calculation time raised about 70 sec, which isn't much compared to the total time of 7343 sec.
So 4.55 seems to be a good fix. :)
cu,
Michael
RE: ... but this also
)
You can use the calibrating client from truXoft:
http://boinc.truxoft.com/core-cal.htm
I compiled a 64-Bit version of it and mailed to truxa, but didn't get an answer so far. If he's interested, this version will be available for download in the future.
cu,
Michael
RE: Waiting for validation
)
OK so far. Result 24636950 for this work unit validated and is the canonical result.
Nothing travels faster than the speed of light with the possible exception of bad news, which obeys its own special laws.
Douglas Adams (1952 - 2001)
I am battling to get 4.55 to
)
I am battling to get 4.55 to run. I just get the following error message.
2006/04/10 13:47:53|Einstein@Home|Unrecoverable error for result z1_0911.5__131_S4R2a_0 (process exited with code 22 (0x16))
I am running TruXoft 5.3.12. I can run the 4.40 workunits fine but as soon as I try the 4.55 it fails immediatelly with the above error.
Anybody know what the error is?
Thx
Richard
Hello RDPearson, I have
)
Hello RDPearson,
I have had the same problem with 4.51. Look at the thread "Linux BETA app 4.51 available" and read the instructions on the beta test app page.
I solved my problem by renaming the .so file and creating a new one but empty.
But please, read the page :
http://einstein.phys.uwm.edu/app_test.php
cu,
Bernard.
RE: I solved my problem by
)
Many thanks Bernard, that fixed it for me too :)
Cheers
Richard
Hi Richard, Glad to hear
)
Hi Richard,
Glad to hear that ;-)
Could you tell us which GNU/Linux distro are you running, which X server, which kernel ?
This way, Bernd can have a better view on which systems 'albert' crashes.
Bernard.