Message boards :
Number crunching :
HADAM3P NOT Compatible with Pentium3s & AthlonXPs
Message board moderation
Author | Message |
---|---|
Send message Joined: 7 Aug 04 Posts: 2168 Credit: 64,545,139 RAC: 6,708 |
hadam3p (including the SAF, EU, PNW variants) is NOT compatible with CPUs/processors without SSE2 capabilities. That means HADAM3P is NOT compatible with
|
Send message Joined: 25 May 06 Posts: 3 Credit: 1,926,186 RAC: 552 |
I am using an Athlon XP 2400 chip and was getting this sort of problem with HADAM3P. I changed my preference to HADSM3 but the problem has not gone away. Program finished after a few minutes without doing any work. Does this problem extend to other models or have I got another problem with similar characteristics. Note: problem does not relate to firewall or antivirus program as I have tried downloading with these turned off. Any help appreciated. thanks |
Send message Joined: 7 Aug 04 Posts: 2168 Credit: 64,545,139 RAC: 6,708 |
All the hadam3p\'s errored almost immediately, as would be expected. The immediate error for the hadsm3 is strange since you have previously run several of these through a phase or two (although none have completed). There has been no change to hadsm3 since Sept of last year, so that is not it. Have you changed anything on your computer lately, hardware or software? Since that PC has trouble completing models, I would run the Prime95 torture test on that PC for at least 12 hours to make sure it is stable. I see you are running boinc 6.4.7. When did you install this? Perhaps try 6.2.19, and if that doesn\'t work, try 5.10.45. You may need to uninstall the current version to reinstall. |
Send message Joined: 27 Jan 07 Posts: 300 Credit: 3,288,263 RAC: 26,370 |
I am using an Athlon XP 2400 chip and was getting this sort of problem with HADAM3P. Those really old systems are susceptible to VCore voltage drops when models are running. I have an old Athlon MP 1200 system that does this. It can run most applications fine, but climate models and other BOINC apps don\'t always finish. By monitoring the voltages, I found the problem. Turns out the only fix is a new motherboard. :( |
Send message Joined: 25 May 06 Posts: 3 Credit: 1,926,186 RAC: 552 |
PC was down for a couple of weeks - but now up and running with new power source and no Lan card (now using wireless), but no other changes in software or hardware. After problems sorted I think boinc was running ok but saw message to upgade to 6.4.7. As soon as upgrade loaded workunit finished as did all subsequent downloads. I tried going back to 6.2.19 but this gave same problem. I then saw this post so I returned to 6.4.7. and tried the hadsm3 workunit which then failed. I have saved the pc messages from the last two failed units (i.e. 6.2.19 with hadam3p and 6.4.7 with hadsm3) - are these any use? I can try the 5.10.45 version but it seems to me that something else is up? Does running part models help? If uncompleted models are no use I will look for shorter workunits or just stick to using my new pc. thanks |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
Partially completed models aren\'t of much use, except if some undergrad is looking at them to see why they failed. The project has moved on from the days when part completed 160 year models were being used to start new, shorter model types. You\'re best bet is the basic slab models, hadsm3. Backups: Here |
Send message Joined: 31 Aug 04 Posts: 42 Credit: 15,303,437 RAC: 130,044 |
Unless the BIOS allows you to set the vcore.I am using an Athlon XP 2400 chip and was getting this sort of problem with HADAM3P. - da shu @ HeliOS, "Free software is a matter of liberty, not price. To understand the concept, you should think of free as in free speech, not as in free beer" |
Send message Joined: 27 Jan 05 Posts: 3 Credit: 918,814 RAC: 115 |
Finally the answer.... This needs an error message, my Athlon XP 3200 has been failing on the CM3P data for weeks, whereas Rosetta has worked perfectly. I checked the FAQs and there was no mention of this problem. How strong is the "prefer this type" preference? |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
The original message in this thread was also posted in the News and Announcements thread, on the same date as here, (29 Mar 2009). The best way to keep up with announcements, is to subscribe to that thread. Backups: Here |
Send message Joined: 7 Aug 04 Posts: 2168 Credit: 64,545,139 RAC: 6,708 |
How strong is the "prefer this type" preference? Not sure what you mean. If you select other model types, and not hadam3p, then you will only download and run the other model types. |
Send message Joined: 27 Jan 05 Posts: 3 Credit: 918,814 RAC: 115 |
How strong is the "prefer this type" preference? Ok, so "Prefer this type" means "choose only this type". Suggest you reword the prompt. prefer: To choose or be in the habit of choosing as more desirable or as having more value which suggests that non preferred options may get chosen occasionally |
Send message Joined: 27 Jan 05 Posts: 3 Credit: 918,814 RAC: 115 |
In order of preference it would be better to: 1. Check the processor capabilities which are known by boinc and not download those units which require capabilities not present 2. Check the capabilities and issue an error message in the boinc messages to the user to change preferences when downloading a unit requiring specific capabilities 3. Run the unit and when erroring, issue a specific error on missing capability 4. Run the unit and when erroring, issue an additional warning to check preferences referencing the issue 5. Provide a link in the boinc messages when writing an error to the common reasons for that error which should include this issue I presume the difficulty is that options 1 & 2 would require a change to Boinc to check the work unit restrictions. |
Send message Joined: 12 Aug 09 Posts: 20 Credit: 3,063,648 RAC: 0 |
I am running an Athlon 64 3500+ which features sse2 but after 112 hours I got this message then the whole thing crashed. Can I turn off the spinning globe thing? Intel® Visual Fortran run-time error Forrtl: severe (38): error during write, unit 8 file C:\\Documents and Settlements\\All Users\\Application Data\\BOINC\\projects\\climateprediction.net\\hadam3p_ma6e_1987_2_1006334352\\tmp\\pipe_dummy
hadam3p_um_6.06_w 0084E86A Unknown Unknown Unknown hadam3p_um_6.06_w 0084AE49 Unknown Unknown Unknown hadam3p_um_6.06_w007D2F4D Unknown Unknown Unknown hadam3p_um_6.06_w 007D2198 Unknown Unknown Unknown hadam3p_um_6.06_w 007AB500 Unknown Unknown Unknown hadam3p_um_6.06_w 00738A6A Unknown Unknown Unknown hadam3p_um_6.06_w 007805C7 Unknown Unknown Unknown kernel32.dll 7C817077 Unknown Unknown Unknown
|
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
To change any graphics option: Click on Show graphics in the Tasks tab of the manager. Press H to get list of options. Press the one that you want to try. To make changes permanent: In Climateprediction.net preferences on Your Account page (blue menu to the left of here), click edit at the bottom, and change the option that you want. Click Update preferences at the bottom. Click the Update button in the Projects tab of the manager, after selecting this project. ****************************** Fortran errors affect some computers some or all of the time. The reason isn\'t known. However, these models are written in Fortran for a supercomputer, and don\'t like being interrupted while running. Some threads discussing it, with suggestions: here, here, here, and here Backups: Here |
©2024 climateprediction.net