Message boards :
Number crunching :
Optimized programs based on SETI 4.18?
Message board moderation
Author | Message |
---|---|
Ron Kosinski Send message Joined: 11 Nov 00 Posts: 42 Credit: 35,074,534 RAC: 22 |
Are there any optimized executibles using the new SETI 4.18 software yet? I am currently using the standard 4.18 software now. I am trying to decide whether to use an optimized 4.11 program now, or wait for one based on 4.18? I am considering either seti-p3.zip from the Marisan website or seti-athlon-xp.zip from the pperry website. System info: AMD Sempron 2800+ CPU-Z info: Name: AMD Athlon MP Thoroughbred Specification: AMD Athlon XP 2400+ Instructions: MMX+, 3DNow+, SSE Windows 2k Pro |
dave015702 Send message Joined: 13 Feb 05 Posts: 271 Credit: 2,341 RAC: 0 |
Are there any optimized executibles using the new SETI 4.18 software yet? I believe that 4.18 is the same as 4.09 except for some fixes to the screensaver graphics. Since the optimized applications don't include the graphics anyway just use the 4.11 optimized application. Wouldn't the YAOSCW-K-r8.1.zip be the right one for you? Help and BOINC documentation is available here. |
Prognatus Send message Joined: 6 Jul 99 Posts: 1600 Credit: 391,546 RAC: 0 |
Ron, I'm running TMR's optimzed 4.11 instead of 4.18 - although all WU's show up as 4.18 anyway. You'll have to change the currently available app_info.xml though. See here for how to do that. |
dave015702 Send message Joined: 13 Feb 05 Posts: 271 Credit: 2,341 RAC: 0 |
And don't forget the boinc-445-sse.zip from the same site so your benchmarks and claimed credit is right. Help and BOINC documentation is available here. |
Ron Kosinski Send message Joined: 11 Nov 00 Posts: 42 Credit: 35,074,534 RAC: 22 |
You'll have to change the currently available app_info.xml though. I can't find the app_info.xml file anwhere on my computer. I am using the new 4.45 version of BOINC Manager, if it makes a difference. |
Ron Kosinski Send message Joined: 11 Nov 00 Posts: 42 Credit: 35,074,534 RAC: 22 |
Wouldn't the YAOSCW-K-r8.1.zip be the right one for you? From what I saw, the K version needs SSE+. According to the CPU Z info, I only have SSE. |
Ron Kosinski Send message Joined: 11 Nov 00 Posts: 42 Credit: 35,074,534 RAC: 22 |
And don't forget the boinc-445-sse.zip from the same site so your benchmarks and claimed credit is right. Should I use the MMX or the SSE version? |
StokeyBob Send message Joined: 31 Aug 03 Posts: 848 Credit: 2,218,691 RAC: 0 |
You'll have to change the currently available app_info.xml though. The app-info.xml file is should be in the YAOSCW-K-r8.1.zip file. I believe the one that is in it was made when we were running a SETI standard 4.09 client. If you don't have any SETI work on your computer that you are upgrading that app-info file should work. If you have work in BOINC and you want to keep it from being deleted you will need to change the 4.09 and 409 to 4.18 and 418 in the app_info.xml file. |
Astro Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
Ron, I believe the K version will work for you. I have seen others with 2800 Semperons using it. I have a 2200+ using it. tony |
Skip Da Shu Send message Joined: 28 Jun 04 Posts: 233 Credit: 431,047 RAC: 0 |
Wouldn't the YAOSCW-K-r8.1.zip be the right one for you? Ron, I think it's the wording on the YAOSCW description... "Intel processors with SSE plus AMD AthlonXP". I think the intent is "Intel with SSE AND AMD AthlonXPs". Not SSE+. It's working on one T-bred XP and several Thorton/Barton XPs here. -- Skip |
Astro Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
@Skip The link to DV is yielding a 404 |
Skip Da Shu Send message Joined: 28 Jun 04 Posts: 233 Credit: 431,047 RAC: 0 |
@Skip Ehkkk... off to investigate... Thank U VERY Much Update: OK, s/b fixed now. The site got a major upgrade last week and a few things got moved around... forgot about the sig line http. - da shu @ HeliOS, "A child's exposure to technology should never be predicated on an ability to afford it." |
Ron Kosinski Send message Joined: 11 Nov 00 Posts: 42 Credit: 35,074,534 RAC: 22 |
Ron, I think it's the wording on the YAOSCW description... "Intel processors with SSE plus AMD AthlonXP". I think the intent is "Intel with SSE AND AMD AthlonXPs". Not SSE+. It's working on one T-bred XP and several Thorton/Barton XPs here. -- Skip I changed both SETI and BOINC Manager programs over to the optimized versions about an hour ago. Everything seems to be working OK. I remember from the last 2 times I changed the BOINC Manager program, my scheduler will go bonkers for a while, which I expected. Thank's everybody for your help and input. |
Skip Da Shu Send message Joined: 28 Jun 04 Posts: 233 Credit: 431,047 RAC: 0 |
BTW, If anybody cares.. After upgrading my v4.45 cc to the appropriate YAOSCW cc my benchmarks rose around 8~9%. I used the K version on all the AMD CPUs. The Thortons/T-Breds like the optimized version just a wee bit more than the Bartons (cache size?). The real winner is the SSE2 P-M and the loser was the PII MMX box. All run WinXP Pro SP2 except the XP1600 runs XP Home and the old PII runs Win98se. CPU, CPU Clock Mhz, % Delta in total of the two benchmarks since last time I saved all the numbers (early May). <OL> <LI>XP 2800 DTR, 2257, +8.56% <LI>XP 2500, 2224, +8.75% <LI>XP 2500 DTR, 2185, +8.75% <LI>XP 2000, 2063, +8.92% <LI>XP 2000, 1999, +12.5%, (previous numbers were a bit low) <LI>XP 2000, 1986, +9.89% <LI>PentM 1.6, 1599, +18.78%, (B version, SSE2) the big winner. <LI>XP 1600, 1400, +8.41% <LI>PII 266, 265, +4.79%, (0 version, MMX) least improved. </OL> -- Skip |
W-K 666 Send message Joined: 18 May 99 Posts: 19063 Credit: 40,757,560 RAC: 67 |
BTW, If anybody cares.. Skip When did you change over to the optimized version on the Pent M, I'm interested cause I go PentM 1.86. At moment with un-optimized I'm doing approx 2 hrs/unit. Andy P.S. I may be away I'm watching Nasa Deep Impacter |
Skip Da Shu Send message Joined: 28 Jun 04 Posts: 233 Credit: 431,047 RAC: 0 |
BTW, If anybody cares.. Just a few hours ago. Only one to check is "Preempted" at 59.18% with 50:49 spent and 0:35 to go. If it's estimate is right that'd put it at 1:26 but I think it'll be more like ~1:45. I'll know tomorrow. Maybe I should add that the FP numbers dropped on all the AMD benchmarks but the integer numbers rose about 5x the drop in floating point. I don't know what the real world effect will be with this. The Pentium-M increased both numbers. The PII had no change in FP and a slight boost in integer. Also found an error in numbers for the XP2800 DTR should be, clock = 2266 and change = +9.22% but I can't edit the original post now. |
Speedy67 & Friends Send message Joined: 14 Jul 99 Posts: 335 Credit: 1,178,138 RAC: 0 |
You'll have to change the currently available app_info.xml though. I updated the files on my site, all optimized clients now include the 4.18 to 4.11 app_info.xml. The 4.09 to 4.11 app_info.xml is still available as a seperate file, just in case anyone needs it. Greetings, Speedy67 |
JohnB175 Send message Joined: 15 Oct 03 Posts: 124 Credit: 321,769 RAC: 0 |
when I moved to the optimized clients I used http://www.marisan.nl/seti/app_info409to411.zip everything has been working smoothly so do I need http://www.marisan.nl/seti/app_info418to411.zip? Does it make any difference in the reporting? Thanks |
StokeyBob Send message Joined: 31 Aug 03 Posts: 848 Credit: 2,218,691 RAC: 0 |
when I moved to the optimized clients I used http://www.marisan.nl/seti/app_info409to411.zip If everything is working fine. No The app_info 409 to 411 was made when we were running the SETI standard client 4.09. It made it possible for the new 4.11 optimized client to run the old work already on our machines marked 4.09. The new app_info 418 to 411 is meant to do the same. When you install it with the Tetsuji optimized 4.11 client it should run all of the work units marked 4.18 that are on your system. |
©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.