Message boards :
Number crunching :
New App May 1?
Message board moderation
Previous · 1 · 2 · 3 · Next
Author | Message |
---|---|
John Clark Send message Joined: 29 Sep 99 Posts: 16515 Credit: 4,418,829 RAC: 0 |
Is this new app for everyone or just the 2 or more cores? I am running SSE2 will it be available for my use? I believe the Alex code port to Windows is only relevant to the Conroe CPUs and up - C2D and C2Qs running SSE3, SSSE3 and SSE4 (Yorkfields). Any older machines with CPUs running MMX, SSE and SSE2 will not benefit. But, I understand the Lunitics Team is looking to port to SSE2 and below. It's good to be back amongst friends and colleagues |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
@Fred, Well My E8400 Wolfdale @3.6 might give you a better indication of what to expect, It's been used for development & long optimisation runs, but WU cpu times should still be indicative. [runs only 32 bit windows most of the time too] Jason "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. |
David Send message Joined: 19 May 99 Posts: 411 Credit: 1,426,457 RAC: 0 |
My host stats for that machine do show me that it reached a gnat's whisker under 2400 on the 8/9 April which would translate to 4800 on 4 cores so Problem is that you dont quite get double when going from 2 to 4 cores using the current app - Both my Q6600 PC's here see a slight improvement in times when dropping 1 or 2 cores out. I think that it may be a memory/cache issue that slows it up slightly and by reducing the load by dropping a core out then it seems to work faster (usually only 5 to 10% decrease in crunch times). Rumour has it that the new app does not suffer from the same limitations (or has less issues), but I'll have to wait and see how it works with my motherboard & chip :) |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
... A tiny correction, any Windows PCs with SSE3 capable CPUs (or Higher), This includes SSE3 capable Intel P4 (Prescott upwards), AMD's (that have SSE3) including Phenom. As they always say ... Your mileage may vary. Jason "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. |
John Clark Send message Joined: 29 Sep 99 Posts: 16515 Credit: 4,418,829 RAC: 0 |
... Thanks Jason - I stand corrected, and the slight widening of the net will be welcome by P4 Prescot ownwers. Pity I cannot make use of it as I have a Prestonia based P4 dual Xeon. I believe someone is running Alex's code on a P4 box with SSE3 from the porting thread? It's good to be back amongst friends and colleagues |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
Yes me.. p4 Cedar Mill (Stock 3.2GHz) "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. |
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
... My opteron 175 says sse3 in cpuid but for some reason I was always told to run the sse2 unless I used the ICC-Patch. Will this change now, will my opty work on sse3 without a patch? Official Abuser of Boinc Buttons... And no good credit hound! |
OzzFan Send message Joined: 9 Apr 02 Posts: 15691 Credit: 84,761,841 RAC: 28 |
... I believe everyone was told that because AMD's implementation of SSE3 wasn't up to par and that SSE2 (on those chips) performed better. It sounds to me that with this new code, even the poorly implemented SSE3 will be faster. If SSE2 is backported with this app, that may change again. |
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
... Well I will check it out and let everyone know...Hopefully this old Opty can do the job...If not it may be time for an upgrade.... Official Abuser of Boinc Buttons... And no good credit hound! |
zoom3+1=4 Send message Joined: 30 Nov 03 Posts: 65765 Credit: 55,293,173 RAC: 49 |
... Hopefully all the Seti users who know where to download the appropriate files, Won't bring down the Lunatics server on Thursday. The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's |
OzzFan Send message Joined: 9 Apr 02 Posts: 15691 Credit: 84,761,841 RAC: 28 |
... I can always host a mirror, but my server capacity isn't that great as I'm limited by Comcast's 1MB upload speed (which would be everyone else's download limitation). I'd still be willing to help out if needed. |
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
... Isn't Thursday just about 7 hours away somewhere? Official Abuser of Boinc Buttons... And no good credit hound! |
Honie Send message Joined: 22 Jan 04 Posts: 141 Credit: 29,681,066 RAC: 0 |
It's out. You can download it now. http://lunatics.kwsn.net/discussion-forum/releases-all-released-versions-are-announced-here.msg7491.html;topicseen#msg7491 |
Dr. C.E.T.I. Send message Joined: 29 Feb 00 Posts: 16019 Credit: 794,685 RAC: 0 |
It's out. You can download it now. thanks Honie - There is no support for CPUs below SSE3 capablity . . .. Windows Port of Alex Kan's v8.0 - Science App Release BOINC Wiki . . . Science Status Page . . . |
zoom3+1=4 Send message Joined: 30 Nov 03 Posts: 65765 Credit: 55,293,173 RAC: 49 |
It's out. You can download it now. @ Honie: Thanks and I've made Your passive link active. The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's |
Honie Send message Joined: 22 Jan 04 Posts: 141 Credit: 29,681,066 RAC: 0 |
It's out. You can download it now. Thanks. Didn't know how to do it. New App is up & running |
Keith T. Send message Joined: 23 Aug 99 Posts: 962 Credit: 537,293 RAC: 9 |
It's out. You can download it now. There are 2 ways to make an active link. [ url ] www.google.com [ /url ] http://www.google.com or [ url= setiathome.berkeley.edu ] Some Text [ /url ] Some Text Take the spaces out. The example in the BBCode instuctions gives the more complex way but not the simple one. Looking foreward to SSE version if it is possible. Sir Arthur C Clarke 1917-2008 |
Robert Curtiss Send message Joined: 9 Apr 08 Posts: 49 Credit: 3,105,422 RAC: 0 |
Before new code average times for me was 3 hours 45 mins for a crunch now the ETA is 10 hours+? |
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
... My opty loves having a 64bit app to run... Official Abuser of Boinc Buttons... And no good credit hound! |
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0 |
Before new code average times for me was 3 hours 45 mins for a crunch now the ETA is 10 hours+? Somehow in the transition you managed to detach and reattach, so your Duration Correction Factor was reset to 1 and is causing the long ETA. It will work down to a realistic value in a few days. Note that if the "instructions.txt" file included in the downloads is followed exactly, no detach and reattach should take place. Joe |
©2024 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.