rpi_logo
App 1.13
App 1.13
log in

Advanced search

Message boards : Number crunching : App 1.13

Author Message
Profile banditwolf
Avatar
Send message
Joined: 12 Nov 07
Posts: 2425
Credit: 524,164
RAC: 0

Message 1031 - Posted: 13 Dec 2007, 3:35:42 UTC
Last modified: 13 Dec 2007, 3:43:26 UTC

I am still getting memory leaks with the 'gs_' wu's. No other problems noticed yet. Almost identical times to complete as the previous wu's.
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

Profile agony
Send message
Joined: 24 Oct 07
Posts: 22
Credit: 130,021
RAC: 0

Message 1035 - Posted: 13 Dec 2007, 8:05:52 UTC

no problems here too since 113 application. what i can see is i beat down the work
~20 secs faster no matter if gs or ps :). happens to both of my computers.

maybe using gcc 4.2x was a good idea ^^
____________

Emanuel
Send message
Joined: 18 Nov 07
Posts: 280
Credit: 2,442,757
RAC: 0

Message 1036 - Posted: 13 Dec 2007, 15:30:46 UTC - in response to Message 1035.

This project is using GCC 4.2? Out of curiosity (sorry if it's off-topic) what optimisations are you using?

As for the new binaries, they did seem a bit faster to me, though I wasn't sure. Crunching is going smoothly :)

C
Send message
Joined: 22 Nov 07
Posts: 36
Credit: 1,224,316
RAC: 0

Message 1039 - Posted: 13 Dec 2007, 20:11:12 UTC

My G5 appears to complete a WU in the same amount of time with either 1.12 or 1.13. My MacBookPro Core Duo, though, takes about 5% less time with 1.13 than it did with 1.12.

No apparent errors or problems with the G5, G4, MGP, or old iMac (333MHz)

C

Profile JLDun
Avatar
Send message
Joined: 17 Nov 07
Posts: 77
Credit: 94,729
RAC: 0

Message 1040 - Posted: 13 Dec 2007, 20:49:15 UTC - in response to Message 1036.

As for the new binaries, they did seem a bit faster to me,

The two mine have done (so far) are about a minute faster (~13:38 vs 14:30-15:10).
____________

Odysseus
Send message
Joined: 10 Nov 07
Posts: 95
Credit: 8,011,810
RAC: 8,734

Message 1041 - Posted: 13 Dec 2007, 20:58:59 UTC - in response to Message 1039.

My G5 appears to complete a WU in the same amount of time with either 1.12 or 1.13. My MacBookPro Core Duo, though, takes about 5% less time with 1.13 than it did with 1.12.

So far my dual G5, running BOINC v5.4.9 under “Tiger”, is looking just about 5% faster with v1.13.

____________

Profile Travis
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 30 Aug 07
Posts: 2046
Credit: 26,480
RAC: 0

Message 1048 - Posted: 14 Dec 2007, 1:15:34 UTC - in response to Message 1041.

My G5 appears to complete a WU in the same amount of time with either 1.12 or 1.13. My MacBookPro Core Duo, though, takes about 5% less time with 1.13 than it did with 1.12.

So far my dual G5, running BOINC v5.4.9 under “Tiger”, is looking just about 5% faster with v1.13.


part of the reason that the work units are running faster is that Nate put in some wrong parameters for the wedge of the sky it's calculating over. we'll be updating the binaries again in another day or two with the new stuff (so times should be back to what they were before). i'll also be starting up some convolution searches soon, and these workunits should take much much longer.


Post to thread

Message boards : Number crunching : App 1.13


Main page · Your account · Message boards


Copyright © 2018 AstroInformatics Group