21)
Message boards :
Number crunching :
4890 support
(Message 28494)
Posted 28 Jul 2009 by speedimic Post: Now then, a little fan and case optimizing and the beast is down to 78C GPU and 76C Mem (running 900/1000). |
22)
Message boards :
Number crunching :
4890 support
(Message 28465)
Posted 27 Jul 2009 by speedimic Post: Just got my 4890 runnin' - the sapphire 900MHz version. Loud, hot and seems fast... check here What do you think - is 90C still ok? [edit] LOL - forgot to plug in one of the case fans... 85C now. [/edit] |
23)
Message boards :
Application Code Discussion :
milkyway & milkywayGPU makefile
(Message 24391)
Posted 6 Jun 2009 by speedimic Post: looks like the cpu is faster... ;)) |
24)
Message boards :
Application Code Discussion :
milkyway & milkywayGPU makefile
(Message 24314)
Posted 5 Jun 2009 by speedimic Post: looks like the cpu is faster... CPU(intel core2 e6750) vs GPU(nv geforce 9600gt) computation times linux_x86_64 on stripe-20. Could you explain this? |
25)
Message boards :
Number crunching :
CPU Work Units Did Not Validate
(Message 22044)
Posted 12 May 2009 by speedimic Post: OK, so even though I can run SSE3 apps on this chip running under WinXP, when I fired up the client running on Ubuntu 9.04 indeed it did not list SSE3 as a feature of my CPU, only SSE2. So now the question is, which app do I want? open a root treminal, type cat /proc/cpuinfoand look if SSE3 will show up for the cpu. The 0.18 mkII marked for Linux 32 bit or the 0.18d (older) marked for Linux 32/64 bit. I'm running the 64 bit OS so as far as I can see it's a choice between one of these two but I'm not sure which one. I didn't make any 64bit SSE2 - assuming that any 64bit capable cpu is SSE3 capable. BTW, mkII or not doesn't really matter. |
26)
Message boards :
Application Code Discussion :
Recompiled Linux 32/64 apps
(Message 17734)
Posted 6 Apr 2009 by speedimic Post: The linux client is simply inefficient compared to win32. To be honest, the win-apps by ClusterPhysik are a bit faster. Those Q6600ers aren't OC'd (the benchmark results might be a little off because of the 6.1 BOINC client) If I recall it right, Cluster made some more modifications which didn't make it into stock code... As stated before my apps are compiled from stock source and approved by Travis. @mfl0p: The compiler flag "-fp-model fast=2" doesn't make any difference - nighter in speed nor in accuracy (at least on my test box). |
27)
Message boards :
Number crunching :
Travis: Please set a minimum update interval
(Message 17556)
Posted 4 Apr 2009 by speedimic Post: ... I don't get it... Can we get money for credits now? |
28)
Message boards :
Number crunching :
Server Updates and Status
(Message 17037)
Posted 27 Mar 2009 by speedimic Post: I'm not sure if we can do this having GPU milkyway as just a separate application or not. That's good question, as BOINC doesn't recognise ATI cards as coprocessors. For the backend a ATI-gpu-cruched Wu must be looking like a cpu-crunched WU... |
29)
Message boards :
Application Code Discussion :
GPU app teaser
(Message 16834)
Posted 25 Mar 2009 by speedimic Post: Just setting up a box for a friend - and could not resist to try his 4850. :)) Well, reading about it is definetly different to seeing it... this IS fast. I gess I got to my one too. Any recommendations? I'm not looking for the fastest, more like "best bang for the buck" |
30)
Message boards :
Number crunching :
Massive team movements
(Message 15511)
Posted 15 Mar 2009 by speedimic Post: looks less strange now... |
31)
Message boards :
Number crunching :
Insta-Purge
(Message 14835)
Posted 10 Mar 2009 by speedimic Post: Travis, looking at my tasklist, I just found 2 Wus reported on the 13th which aren't insta-purged... Whatever "blarg" is - it didn't help. ;)) |
32)
Message boards :
Number crunching :
ScreenSaver?
(Message 14833)
Posted 10 Mar 2009 by speedimic Post: Besides, why do we care *WHY* some one does BOINC ... the end is the same ... yep, tried a comeback 2 weeks ago... half the wus didn't even dl and the forum looks like project staff has forgotten where it is... |
33)
Message boards :
Number crunching :
ScreenSaver?
(Message 14827)
Posted 10 Mar 2009 by speedimic Post: Besides, why do we care *WHY* some one does BOINC ... the end is the same ... Cosmo? ;) |
34)
Message boards :
Application Code Discussion :
milkyway released under GPLv3
(Message 14826)
Posted 10 Mar 2009 by speedimic Post: Talking about copyright, ist this GPL? numericalIntegration.c: /*Gauss-Legendre quadrature taken from Numerical Recipes in C*/ |
35)
Message boards :
Number crunching :
Milestones
(Message 14825)
Posted 10 Mar 2009 by speedimic Post: let's see... MW close to 3.000.000 BOINC overall a good 13.000.000 [edit]damn, I need a GPU...[/edit] |
36)
Message boards :
Number crunching :
zsling.com - BOINC links
(Message 14428)
Posted 8 Mar 2009 by speedimic Post:
no, I won't see the planes then. ;) Just installed Opera mini - by far not as pretty and handy as Chrome - but I can read the posts. Put in your sig whatever you like :P |
37)
Message boards :
Number crunching :
zsling.com - BOINC links
(Message 14368)
Posted 8 Mar 2009 by speedimic Post: ... hmm... won't improve the situation for me. :( I'll try to get another browser... |
38)
Message boards :
Cafe MilkyWay :
Word Link
(Message 14256)
Posted 7 Mar 2009 by speedimic Post: aloha |
39)
Message boards :
Number crunching :
Coprocessor info in xml stats
(Message 14245)
Posted 7 Mar 2009 by speedimic Post: Are you planning on offering the coproc information in the xml stats at all ? ...or have separate stats for GPU like suggested here: Separated stats for CPUs and GPUs. |
40)
Message boards :
Number crunching :
zsling.com - BOINC links
(Message 14226)
Posted 7 Mar 2009 by speedimic Post: ... LOL, now I know why the browser of my G1 freaks out when it comes showing your posts. Must be that animated stuff... |
©2024 Astroinformatics Group