Message boards :
Number crunching :
Optimized Astropulse for Windows issues, install, questions
Message board moderation
Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · 8 · 9 . . . 12 · Next
Author | Message |
---|---|
Bob Mahoney Design Send message Joined: 4 Apr 04 Posts: 178 Credit: 9,205,632 RAC: 0 |
Checking your result it shows you are still running the stock on those AP units. Acknowledged. Note to everyone: My previous message about 8200RAC becoming 3400RAC with the switch to Astropulse is wrong, wrong, wrong. I want to see Astropulse thrive, and such messages might give the wrong impression. I messed up the installation of the new application; that was the only problem. Bob |
Fred W Send message Joined: 13 Jun 99 Posts: 2524 Credit: 11,954,210 RAC: 0 |
[@Bob, Must admit, I think I've got to the max on pneumatic cooling - and reasonably quiet as the case fans run at 75% most of the time. But I have been thinking about water as my son-in-law has a Q6600 rig with dual Reserator cooling (only one I have come across) and that is "quiet"!! F. |
bounty.hunter Send message Joined: 22 Mar 04 Posts: 442 Credit: 459,063 RAC: 0 |
Optimized installed and running fine. Althought the first one I've sent in has validated properly, it has underclaimed... It claimed 753.35 as against 771.52 by the other 2 wingmen. http://setiathome.berkeley.edu/workunit.php?wuid=329831449 Here is the result......http://setiathome.berkeley.edu/result.php?resultid=1016587852 Problems with the credit multiplier ? |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
It looks like I'm on target for 60 hours -- compared to 150 hours on the non-optimized AP. This on a machine that does not benefit from the current optimized Multibeam applications. |
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0 |
Optimized installed and running fine. No, it's just that your unit was sent on October 10, so the server-side credit multiplier for that day was applied to your result. The other two were sent September 10 so had the credit multiplier from that date. The differential can go either way, in this case it worked to your advantage. I am a bit surprised it's still varying that much, but considering how long AP work takes on some systems I suppose a long settling time was inevitable. Joe |
HFB1217 Send message Joined: 25 Dec 05 Posts: 102 Credit: 9,424,572 RAC: 0 |
Just a it of an update my Q9550 with an OC of 3.86gig is completing AP work units in just about 12 hours plus or minus about 8 minutes. All my AP units are completing sucessfully and the ones not pending have been validated and awarded about 750+ credits. It also seems to be that I am only getting AP work units no problem with that for me and I am in the process of replacing another Pent D with a duplicate Q9550 system. The build is under way and soon to be runing PI 95 testing prior to allowing Seti to run on it. Come and Visit Us at BBR TeamStarFire ****My 9th year of Seti****A Founding Member of the Original Seti Team Starfire at Broadband Reports.com **** |
KenZaske Send message Joined: 12 Oct 04 Posts: 7 Credit: 456,380 RAC: 0 |
If you can provide the parameters, at the very least I can build an old style batch file. An MSI installer should be able to determin where to put the files and append the correct text to the app_info.xml file. If necessary I will get a friend to help me (he codes installers for HP Lazer Jet) I am sure we can figure out something. Perhaps even include MB with AP. When it comes to BOINC I am a total noob, but batch files and msi files I can almost write in my sleep. |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
If you can provide the parameters, at the very least I can build an old style batch file. An MSI installer should be able to determin where to put the files and append the correct text to the app_info.xml file. If necessary I will get a friend to help me (he codes installers for HP Lazer Jet) I am sure we can figure out something. Perhaps even include MB with AP. Much appreciated. It's something none of us have had the time to look at more than cursorily. To clarify, the complexity lies not so much with copying the files as such, or stopping and restarting Boinc during the process, but with locating the project folder. I'd imagine checking some key locations for the setiathome.berkeley.edu project folder would be quite feasible, as would possibly selecting/detecting the relevant applications from those available, and generating the appropriate app_info.xml file. If you feel this is workable, given your knowledge of msi files, feel free to drop in at Lunatics. if you have the time you can register there and we can get a discussion going, otherwise PM here could work too, though doesn't have the advantage of involving the other developers. I could probably modify some cpuid code I have to provide SSE level / brand infomation from a stub exe you could parse to help select defaults (preferably optionally overridable by the user). "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. |
Mumps [MM] Send message Joined: 11 Feb 08 Posts: 4454 Credit: 100,893,853 RAC: 30 |
I wander if anyone, has tested it on XP64, running a AK_V8_WINX64_SSSE3x and the new 32BIT AP optimized app. Well, I've now completed two AP's using the Optimized App. Both completed successfully, whereas all 5 WU's this machine crunched using the stock app errored out well before completion. Once again, great work guys! PS: About 16 hours each, versus the longest previous one hitting 30 hours before crapping out... |
Fred J. Verster Send message Joined: 21 Apr 04 Posts: 3252 Credit: 31,903,643 RAC: 0 |
Hi, Crunchers ;) , @ Raistmer and Jason, thank you for your explanation. @ Mumps, yeah, lately the errors on EINSTEIN, are now maybe 1 in a >250. Bought OCZ reaper(6500) runs now @ 420MHz a stick (FSB/RAM=5:6) and 2.1V. With 5.5.5.18 timings. Gonna try the optimized AP app. at first on my Q6600(@3GHz), WIN XP32BIT and 2 C2D's Is there any noticeble difference between the 5.10.xx & 6.2.18 and up, builds? Keep On Crunching |
Fred W Send message Joined: 13 Jun 99 Posts: 2524 Credit: 11,954,210 RAC: 0 |
Hi, Crunchers ;) , @ Raistmer and Jason, thank you for your explanation. 5.10.xx / 6.2.xx won't affect crunching speed. I'm still running 5.10.45 on my Vista64 rig as so many people seem to be having problems with 6.2.xx. F. |
lemonade lucy Send message Joined: 23 Sep 99 Posts: 73 Credit: 1,454,014 RAC: 0 |
early in this thread a note makes it clear that the app_info.xml must be edited or created. understood. my question: what's the trigger for file creation, pre opti edit? that is, it seems many are not having to create the file, that it created as result of whatever config. did I miss something? if this addressed in another thread, please pass along the link. thanks. |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14656 Credit: 200,643,578 RAC: 874 |
early in this thread a note makes it clear that the app_info.xml must be edited or created. understood. my question: what's the trigger for file creation, pre opti edit? that is, it seems many are not having to create the file, that it created as result of whatever config. did I miss something? To start with, there was just an optimised application for the standard SETI ('Multibeam') search. To make things easy, the optimisers included a suitable app_info.xml in the MB download packages. Most people got their first one that way. The new Astropulse optimised packages also contain an app_info.xml file, and that's all you need if you only want to run the Astropulse search - it's self-contained. The problem arises if you want to run both searches on a single machine. You can only have one app_info.xml file, so you have to combine the active ingredients of both pro-forma, pre-supplied files. If you do want to run both types of application, and so have to edit the two halves together, make sure you use a plain-text editor (Notepad or Wordpad) throughout. People have come to grief opening one part in Internet Explorer and copying from that - IE messes up the formatting. |
Byron S Goodgame Send message Joined: 16 Jan 06 Posts: 1145 Credit: 3,936,993 RAC: 0 |
|
Cosmic_Ocean Send message Joined: 23 Dec 00 Posts: 3027 Credit: 13,516,867 RAC: 13 |
I would love to give this a shot on my Opteron 2210's, buuut... I've been looking for 20 minutes now and I don't appear to have an app_info.xml..at all. Searched the entire system, including hidden and system folders. Using BOINC 6.2.19, stock everything except my cc_config.xml file. Any suggestions? Only thing I can think of is out in the main BOINC directory, client_state.xml has file info including byte count and a hash. Linux laptop: record uptime: 1511d 20h 19m (ended due to the power brick giving-up) |
Leaps-from-Shadows Send message Joined: 11 Aug 08 Posts: 323 Credit: 259,220 RAC: 0 |
You appear to be running the standard applications for both Enhanced and Astropulse units. If that's the case, you don't have an app_info.xml file on your system - you don't actually need the file until you use an optimized app. I highly recommend following the instructions here, with one modification. The only thing you need to change is to download this file instead of the first one arkayn suggests to Leo. Why? The first one arkayn suggests is Intel-only. You'll still need to download the second file arkayn suggests - that's the optimized Astropulse file. Make sure you edit the included app_info.xml file with Notepad or Wordpad. IE screws things up, as Leo found out. You won't regret running the optimized apps - they reduce completion times significantly for both Multibeam and Astropulse units. On my primary machine (quad-core 2.3GHz Phenom), shorty Multibeams finish in around 00:34:00, regular Multibeams in around 01:24:00, and Astropulse units in around 28:40:00. That's a huge improvement - especially for the Astropulse units. 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 |
Cosmic_Ocean Send message Joined: 23 Dec 00 Posts: 3027 Credit: 13,516,867 RAC: 13 |
Alright. I kind of made an oops. I was unaware that MB and AP had to be in app_info, so it dumped my 60-ish MB WUs since it couldn't find an application to run them. Lesson learned. So far, I notice the 3 APs that were in progress with the stock app are now going with some decent percentage jumps, and the ETA is counting down about 8 seconds per update interval. The one that just started is still saying 123h and is counting down about 1 second per interval. Should speed up as it goes. [update: after 30 minutes, the in-progress APs dropped back to 1-second intervals with still 60-something hours remaining. I should have just left well enough alone.] Linux laptop: record uptime: 1511d 20h 19m (ended due to the power brick giving-up) |
Byron S Goodgame Send message Joined: 16 Jan 06 Posts: 1145 Credit: 3,936,993 RAC: 0 |
|
[B^S] madmac Send message Joined: 9 Feb 04 Posts: 1175 Credit: 4,754,897 RAC: 0 |
|
Leaps-from-Shadows Send message Joined: 11 Aug 08 Posts: 323 Credit: 259,220 RAC: 0 |
That's why I said to follow those instructions - they covered both Multibeam and Astropulse. I should have just left well enough alone. Oh come on. Really? You're already regretting it, without even giving it a chance? You really will see big improvements in completion times. I got about 33% improvement in Multibeam, and 70.5% improvement in Astropulse. Running optimized apps for both is definitely worth it. 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 |
©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.