Message boards :
Number crunching :
Optimized Astropulse for Windows issues, install, questions
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 . . . 12 · Next
Author | Message |
---|---|
dnolan Send message Joined: 30 Aug 01 Posts: 1228 Credit: 47,779,411 RAC: 32 |
Try winrar, works great for me. Google it if you're interested. -Dave |
Leaps-from-Shadows Send message Joined: 11 Aug 08 Posts: 323 Credit: 259,220 RAC: 0 |
Okay. It's official - I finally have one all-optimized Astropulse unit finished. The completion time was 28:49:52. I've also got another one crunching now at about 40% done - looks like it'll finish in around the same time as the first one. That's a 70.5% improvement with the optimized SSE3 application on my Phenom 9650 (quad-core, 2.3GHz). I can't fully express how awesome it is to be able to crunch Astropulse units in 1.20 days instead of 4.08 days with the standard app. I've had a couple of partially-optimized Astropulse units validate already, so it shouldn't be a problem on the all-optimized units. In case it's not clear, let me say: Thank you, thank you, thank you, thank you, thank you ... to all who were involved in bringing us this miracle application. *bows and scrapes* We're not worthy! Cruiser Gateway GT5692 L-f-S Edition -Phenom X4 9650 CPU -4GB 667MHz DDR2 RAM -500GB SATA HD -Vista x64 SP1 -BOINC 6.2.19 32-bit client -SSE3 optimized 32-bit apps |
Andy Lee Robinson Send message Joined: 8 Dec 05 Posts: 630 Credit: 59,973,836 RAC: 0 |
I've just edited the app_info and removed the offending spaces. I checked it by opening it with IE (which shows it all nicely formatted) and it has added the space before the slash for the "main program" and "executable". Something else we can blame Microsoft for. No, not Microsoft this time. Blame Netscape! <tag /> became standard for end tags because of an old bug in Netscape. If you look at any xhtml code you will always see <br /> instead of <br/>. The space before the slash should never ever break any xml code. Andy. |
arkayn Send message Joined: 14 May 99 Posts: 4438 Credit: 55,006,323 RAC: 0 |
|
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
No, not Microsoft this time. Blame Netscape! LoL, that's most likely. Unfortunately, we're pretty much stuck with whatever format boincapi/seti_boinc parsing routines are implemented as stock, for reasons of state, init, config & log file compatibility), rather than correcting the processing to follow standards & cope with quirks. Of course, rewriting a goodly portion of boincapi & seti_boinc to use a more bullet proof 3rd party xml parsing library is an option for the ultimate masochist, but likely render a new application incompatible with stock applications and perhaps boinc version dependant. 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. |
MarkJ Send message Joined: 17 Feb 08 Posts: 1139 Credit: 80,854,192 RAC: 5 |
Thanks, Mark, that explains the conundrum perfectly - I've posted the solution to my bug-reporting thread app_info.xml parsing problem at BOINC Development. One of my machines has managed to pickup a couple of AP work units and its crunching one as I type this, using the opti app so that seems to have done the trick. I'll have to run around to the other machines and upgrade the rest of them. Einstein has also got a optimised SSE2 app so I need to put it on them as well. Oh and the uploads seem to be on the blink so I can't send the Seti results back for the stuff that has processed. BOINC blog |
OzzFan Send message Joined: 9 Apr 02 Posts: 15691 Credit: 84,761,841 RAC: 28 |
I refuse to give in. I will not install 7zip or any other zip like application. I would love to try this out but the format is not a valid zip file. Please post a real zip file. As DNolan suggested, WinRAR works very well and is compatible with 7Zip compression. In fact, that's what I used to unzip the original apps and re-zipped them into .ZIP format. I'm not certain a self-installing .MSI package can merge the contents of the app_info.xml file properly in all cases. Could be too dangerous to do that if it doesn't work properly. Its probably best just to leave it as is. |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
Would you like me to convert it to a self installing MSI? 1) 7z format was chosen from RAR (most conventional format for me cause it has all abilities I need and fully integrates with my favorite file manager FAR), ZIP and 7Z. Criterion was smallest archive size for decreasing bandwidth consuming. 7Z gave smallest archive so it was used. This app is intended for advanced users. So ability to handle different archive formates was supposed. I posted link to archive manager that handles all 3 formates listed in this message. The same could be done with shareware WinRAR application (http://www.rarlab.com). 7z-format was used for Lunatics releases before and could be used again in future, just keep this fact in mind when deciding what archive manager to use. 2) Self-installing to where? There are many places where project directory could be on user PC, there are many combinations of MB app it can use. If all these situations could be handled by single self-installing archive it would be great addition... |
[B^S] madmac Send message Joined: 9 Feb 04 Posts: 1175 Credit: 4,754,897 RAC: 0 |
|
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
I have updated seems to be more heat coming out of my machine however many thanks to Logan for his help. Is there something I can do to turn this heat down? LoL, ultimate solution is to switch PC down ;) If CPU works then it generates heat. If CPU works more intensive then it generates more heat. If you use optimized app then CPU works more intensive hence it generates more heat.... But lets consider not CPU-only power but total power of whole system. Assume stock AP would be finished in X hours. That is it takes X*(Power of CPU + Power of rest system ) kW*h of energy to complete. And now opt app comes in play. Asume that it completes AP task for X/2 hours (true assumption for current release). 1) CPU power will increase but not twice (!). 2) Power of rest system will increase in much lower degree (sure much lower than twice). In first approach it could be considered as constant. Now what we will have: Energy needed to complete the same task will be: X/2*(Power of CPU*y+Power of rest system) where y<2. Evidently, we will get value much lower than we got for stock AP. That is, although CPU power consumption is increased indeed, opt app much more "energy-friendly". So, by using opt app you are helping to save nature ;) . |
[B^S] madmac Send message Joined: 9 Feb 04 Posts: 1175 Credit: 4,754,897 RAC: 0 |
|
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
That a p4? My p4 3.2 runs 24x7 with a reduced fan speed at 48 to 52 degrees C, stock cooler, controlled by speedfan running as a service to maintain 50C. 18 months thereabouts with no downtime apart from windows updates & dust cleanouts, so I'd say you're good to go, excepted any different special environmental considerations you might have. "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. |
[B^S] madmac Send message Joined: 9 Feb 04 Posts: 1175 Credit: 4,754,897 RAC: 0 |
That a p4? My p4 3.2 runs 24x7 with a reduced fan speed at 48 to 52 degrees C, stock cooler, controlled by speedfan running as a service to maintain 50C. 18 months thereabouts with no downtime apart from windows updates & dust cleanouts, so I'd say you're good to go, excepted any different special environmental considerations you might have. No, there is hardly any movement of air where I have my computer, I suppose it needs a good cleanout which I will do at the end of the week. |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
Yes that helps a lot. A friend of mine owned a Prescott ("PressHot")p4, which he didn't clean out for 2 years! I helped him remove the thick dust blanket covering air intakes, removed cleaned & reseated the heatsink with new paste and he was amazed that idle temps dropped from 65 degrees down to about 30, and he could then open multiple browser windows again without the CPU 'freezing' :D Good luck. "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. |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14656 Credit: 200,643,578 RAC: 874 |
I had to clean out one for a neighbour where it was getting so hot that the whole machine shut itself down ~10 minutes after switchon. That CPU was the P4 generation, but branded Celeron - possibly even Celeron D. The stock cooler was mounted on top of a finned heatsink, and blew air downwards into the heatsink. The problem was a layer of dust between the fan and the top surface of the heatsink, which had practically turned into felt and completely blocked the air flow. I think it would have been sufficient just to remove the felt layer: following the advice of crunchers here, I did remove, clean and re-apply the thermal paste too, but you could probably leave that step out (at least initially, to see how it goes) - saves a shopping trip if you don't have the materials to hand. PS I did ask, and there were no cats in that neighbour's household - so it isn't just kitty fur that causes these problems! |
Bob Mahoney Design Send message Joined: 4 Apr 04 Posts: 178 Credit: 9,205,632 RAC: 0 |
I have a question, but first some info . . . My fastest AP WU completion: Time for AP WU: 21h 22m (76,934sec) Credit for AP WU: 752.8 Host- Intel 45nm quad running at 4.2GHz, memory read access at 10,675MB/s The AP WU result has already been validated, so it ran accurately. In a perfect world (no bad results, etc.) this gives a RAC of: 86,400seconds / 76,934seconds x 4cores x 752.8credit = 3,382RAC This machine was consistently running at an RAC rate of around 8,200 for the past couple weeks with optimized Multibeam. Question 1: Does 21hours sound right, or am I accidentally running a NON-optimized Astropulse application? Question 2: Does it sound right that this machine will drop from 8,nnnRAC to 3,nnnRAC if I stick exclusively with Astropulse? Thanks in advance, Bob |
[B^S] madmac Send message Joined: 9 Feb 04 Posts: 1175 Credit: 4,754,897 RAC: 0 |
That a p4? My p4 3.2 runs 24x7 with a reduced fan speed at 48 to 52 degrees C, stock cooler, controlled by speedfan running as a service to maintain 50C. 18 months thereabouts with no downtime apart from windows updates & dust cleanouts, so I'd say you're good to go, excepted any different special environmental considerations you might have. Clean it out today see other topic, temp now down to around 34C, it took me an hour to clean it out. |
Fred W Send message Joined: 13 Jun 99 Posts: 2524 Credit: 11,954,210 RAC: 0 |
I have a question, but first some info . . . @Bob, My Q9450 is o/c'd only to 3560MHz and does an AP WU in a smidgeon under 13 hours with the Optimised App compared to just over 26 hours with the Stock AP App. With an o/c over 4GHz I guess you are on a liquid diet. Looks to me like you are still running Stock with those times. Windows Task Manager will tell you next time an AP WU comes to the top of your heap. F. |
arkayn Send message Joined: 14 May 99 Posts: 4438 Credit: 55,006,323 RAC: 0 |
|
Bob Mahoney Design Send message Joined: 4 Apr 04 Posts: 178 Credit: 9,205,632 RAC: 0 |
[@Bob, Either you own The Mother of All Q9450's, or I messed up the optimized AP app installation on all five of my hosts yesterday. The latter is probably true. The former may still be true! Yeah, the host in question is water cooled with a Zalman Reserator XT. Easy install, simple setup, quiet. Yet it seems so terribly wrong to pipe water into and around naked electronic components. Bob |
©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.