Welcome to MilkyWay@home

Posts by kimmerin

1) Message boards : Number crunching : Future ATI Card Support? (Message 25551)
Posted 15 Jun 2009 by kimmerin
Post:
Just to add my two cents. I most likely will stick with this project rather than switching to the GPU-only project. Why? Because I assume that the same will happen as happened e.g. with Folding@Home or GPUGrid: The WUs will be too big to be processed within the deadline.

I switch off the computer at night, so the crunching really only takes place while I'm actually use my computer for not playing games. So often enough the deadline e.g. for a WU of Folding@Home is reached before this huge thing is calculated through. The CPU-based WUs of M@H are perfect for this type of usage. With the bigger sizes that are currently delivered, one WU takes between 30 seconds up to one minute. So you can easily wait for the end of one calculation-round before shutting down the computer.


Regards, Lothar
2) Message boards : Number crunching : this site has been slow (Message 25178)
Posted 12 Jun 2009 by kimmerin
Post:
Wouldn't help everything, because the website still uses the database used by BOINC. What would probably be better is to split the DB server off.


Set up a master-slave-system and let the webserver take the data from the slave.
That way you can reduce the load to the main server and you have a backup in case the server crashes (again).


Regards, Lothar
3) Message boards : Number crunching : Why is this so? (Message 17970)
Posted 8 Apr 2009 by kimmerin
Post:
I'm not sure why it has changed after the outage, perhaps the database inconsistency slowed things up.


Is it a MySQL? This can be. We have had strange effects as well after a database-crash. As well, it's never wrong to do an occasional "optimize table ..." (we do it every night by a cronjob since then).


Regards, Lothar
4) Message boards : Number crunching : Run the project only a few hours a day (Message 17965)
Posted 8 Apr 2009 by kimmerin
Post:

  • I am entitled to a fair days' credit for a fair days' work
  • I invested substantially in kit to advance science
  • I am entitled to see all my rigs are fully fed with work
  • Button-mashers, script-kiddies and credit-whores are trying to stop me and sabotage the project.




I don’t know Phil. I see BOINC as a 'hobby' and wouldn't spend more on it than I would on the odd computer game/other software now and then.


Some have personal reasons, though.


Regards, Lothar
5) Message boards : Number crunching : Travis: Please set a minimum update interval (Message 17583)
Posted 4 Apr 2009 by kimmerin
Post:
This computer is used for an office automation use and for some little games as Fifa 08 so ...


So as long as you're not do games testing as a living, we can conclude that you bought your graphic card for pure fun.

I don't know why I tell you this, I do what I want with my money that's all. A lot of people do same thing and so what? It's our hobby ! Can you Respect this?


So you buy graphic cards as a hobby without having fun? So if your hobby is computer games and you are still unhappy about it, you should consider changing your hobby.

I bought my graphic card for the same reason (computer games) and am happy about the increase of details games can now produce. And I'm also happy that I can contribute this huge amount of calculation power to a project where one single graphic card is crunching double the amount of WUs than all the computers I combined together to crunch Einstein@Home-WUs (it dropped from seven times when the server was able to deliver more WUs).

Am I unhappy about that? No. I can see progress here being done and it's just a matter of time when other projects will also take advantage of this technique. So all the upset of people with GPUs claiming their first right on WU-reception is just a matter of time and you can continue be unhappy with your hobby as you wish.


Regards, Lothar
6) Message boards : Number crunching : Travis: Please set a minimum update interval (Message 17546)
Posted 4 Apr 2009 by kimmerin
Post:
I don't buy graphics cards for fun !


No? So you run CAD-applications on your Celeron-based box?

Sorry, but buying new equipment for BOINC only is in my eyes contradicting the idea of using "unused" cycles.

Save your money for something else, the project itself will most likely not finish much faster with your contribution.


Regards, Lothar
7) Message boards : Number crunching : Bit of advise (Message 17076)
Posted 28 Mar 2009 by kimmerin
Post:
Yes labstaff misunderstood when i said create an apache alias to milkyway_gpu. They thought I meant delete milkyway and make milkyway_gpu. Looks like things are back though. Sorry about that.

Did you tell something as well to the labstaff of Einstein@Home? Their site is away as well:

Einstein@Home is down due to a fileserver crash. We are working to bring the project back online ASAP. UPDATE: Fri Mar 27 22:39:13 UTC 2009 The filesystem needs to be repaired so the project will likely be down for at least 12 more hours. Thank you for your patience.

Murphy seems to do overtime hours at the moment.


Regards, Lothar
8) Message boards : Number crunching : WU abuse (Message 16402)
Posted 21 Mar 2009 by kimmerin
Post:
Still trying to defend your update script? Tell that to the forum administrator, Dave, who just said that the server can't keep up with the ATI cards.


Dave: But I'm almost certain that the server can't keep up with all the requests from the users with graphics cards.


Oh how often I was "almost certain" just to be proven wrong afterwards ;-) I'm sure they will take care of, because I also think that they have interest in faster WU-computation and don't like if the computation stuck just because their server is not able to deliver fast enough.


Regards, Lothar
9) Message boards : Application Code Discussion : ATI-optimized app 0.19e crashing on XP32 (Message 16234)
Posted 20 Mar 2009 by kimmerin
Post:
8.12 seems to work (9.1 is not available at the sapphiretech-website, so I can't try 9.1) Is there a specific reason why this quite old driver works and newer ones not?

The reason is because with versions 9.2 and up, AMD/ATI renamed some of the .dlls required for the MW app to work. Previous versions have the old filenames, so they will work.


Copying the DLLs with the old name let the computation start, but as I mentioned, the VPU Recover immediately recovers because the driver is regarded as being unresponsive. This stops the calculation and doesn't happen with 8.12.. I just wonder why that is.


Regards, Lothar
10) Message boards : Application Code Discussion : ATI-optimized app 0.19e crashing on XP32 (Message 16228)
Posted 20 Mar 2009 by kimmerin
Post:
The versions 8.12 and 9.1 are supposed to work (on some systems WinXP systems only 8.12 appears to work, don't know why) without any modification. That are the versions you may try.


8.12 seems to work (9.1 is not available at the sapphiretech-website, so I can't try 9.1) Is there a specific reason why this quite old driver works and newer ones not?


Regards, Lothar
11) Message boards : Application Code Discussion : ATI-optimized app 0.19e crashing on XP32 (Message 16050)
Posted 19 Mar 2009 by kimmerin
Post:
Have you tried another driver or reinstalling it?


Before 8.9 was installed, I installed 9.2 just today, about 20 minutes before I posted the first message.


Regards, Lothar
12) Message boards : Application Code Discussion : ATI-optimized app 0.19e crashing on XP32 (Message 16049)
Posted 19 Mar 2009 by kimmerin
Post:
ATI Radeon HD 4870 (1GByte)
Catalyst 9.2


NOTE: ATI Catalyst 9.3 does not work properly on Windows XP systems for ATI CAL applications, including Brook+ applications. If you are using a Windows XP system, you will need to use ATI Catalyst 9.2 or wait for a later driver.


So is 9.2 the new 9.3 then? ;-)


Regards, Lothar
13) Message boards : Application Code Discussion : ATI-optimized app 0.19e crashing on XP32 (Message 16043)
Posted 19 Mar 2009 by kimmerin
Post:
ATI Radeon HD 4870 (1GByte)
Catalyst 9.2

A you don't say anything about it I suppose you did not read the instructions for getting Cat 9.2 (applies to 9.3 too) to work with the MW GPU app.


Oh, there was something after the Change-Log. I stopped reading when reaching there. Maybe the "problems-section" should be placed before the changes-log.

You need to copy and rename some dll files as described in the supplied readme file.


I've done that and after restarting the client, the VPU-recover immediately kicks in. After stopping CCC (including the kill of the process using Sysinternal's ProcessExplorer) I get the same result. I tried killing all processes belonging to ATI but the only effect is that the recover doesn't work and the screen gets dark, forcing a system-restart.


Regards, Lothar
14) Message boards : Application Code Discussion : ATI-optimized app 0.19e crashing on XP32 (Message 16040)
Posted 19 Mar 2009 by kimmerin
Post:
Well you have a valid result now.


I use the SSSE3-optimized app. That only proves that I'm able to install optional applications but still leaves the graphical adapter unused.


Regards, Lothar
15) Message boards : Application Code Discussion : ATI-optimized app 0.19e crashing on XP32 (Message 16036)
Posted 19 Mar 2009 by kimmerin
Post:
When you added the opti app did you turn boinc off? The ones in your que will error out usually.


As you can see in the list of WUs for that system, I downloaded them yesterday. Because I haven't had installed the current version of Catalyse I downloaded 9.2 today and installed it. So I don't know if BOINC was down when I copied the app into the folder, but since then I even rebooted the system, so the crashed ones that you can see in the list right now are not because of BOINC running while copying the app.

Others should come along and give better help.


I hope. Having this mini-supercomputer residing in my box without doing anything than showing the standard-desktop of XP nags me ;-)


Regards, Lothar
16) Message boards : Application Code Discussion : ATI-optimized app 0.19e crashing on XP32 (Message 16033)
Posted 19 Mar 2009 by kimmerin
Post:
Hello,

trying to benefit from the ATI-versions I tried the version 0.19e that is available at zslip. I copied the three files (exe, dll, and app_info.xml) into the project-folder of M@H. Since that time all WUs are crashing immediately as you can see at http://milkyway.cs.rpi.edu/milkyway/results.php?hostid=56103

Something that I can come up with to help finding the reason for this?

Here are the system-specs:

Windows XP 32 Bit on a Dual Core System
ATI Radeon HD 4870 (1GByte)
Catalyst 9.2

The STDOUT of the processes is not very informative (at least for me):

<core_client_version>6.2.19</core_client_version>
<![CDATA[
<message>
- exit code -1073741515 (0xc0000135)
</message>
]]>



Regards, Lothar




©2021 Astroinformatics Group