Welcome to MilkyWay@home

Posts by Slicker [TopGun]

21) Message boards : Number crunching : Is anyone glad with the project? (Message 18985)
Posted 16 Apr 2009 by Profile Slicker [TopGun]
Post:
The people who have had their accounts deleted for no reason yesterday are very, very, very unhappy with this project. I know at least one of those people was a top 10 participant. I really hope the reasone we've heard so little about this "problem" is that Travis is working so hard to make this right.
22) Message boards : Number crunching : BOINC debt - long & short term - what does it do? (Message 17539)
Posted 4 Apr 2009 by Profile Slicker [TopGun]
Post:
Rather than edit the xml by hand, you can use the set_debts command from the boinccmd.exe. Then, to set the debt, you only need to pass in 0.0 instead of 0.000000 for each number. Use "boinccmd.exe --help" to see the syntax for all the commands.
23) Message boards : Application Code Discussion : GPU app teaser (Message 16010)
Posted 19 Mar 2009 by Profile Slicker [TopGun]
Post:
One more question - how do you prevent other project to soak all cache for WUs? I want to crunch Einstein on CPU along with Milky on GPU, but Milky is limited for 6 WUs/CPU, then Einstein fills up the cache (I have it set for 1 day) and then Milky does not request more work. I had to put No New Work for Einstein and suspend couple of its WUs to keep GPU busy.


One word: VirtualBox
24) Message boards : Number crunching : 1 Credit per Second per GPU? (Message 16009)
Posted 19 Mar 2009 by Profile Slicker [TopGun]
Post:
When the GPU client first came out, I think I hit some lucky WUs. I was averaging 1.21 C/S. Now, It's around 0.96.
This is on a non-overclocked 4850.


How do you have yours set up? And which OS, etc? I have a standard clock 4850 too, but it still runs about 35 seconds per 2 WUs(simultaneous)... which now that I think about it is pretty good C/S... maybe is just how often it has no work.


Try the latest 0.19e client which uses much less cpu time. 1 credit per 1-2 cpu seconds is pretty doable with a 48xx card running 0.19e. Granted, that's not clock time or gpu time. The older ATI versions used much more cpu time to wait for the gpu to get done. Kudos to C.P. on the oustanding work done to develop the ATI apps and use so little cpu time.
25) Message boards : Number crunching : Heart of Gold (Message 16008)
Posted 19 Mar 2009 by Profile Slicker [TopGun]
Post:
Some say that credits are valueless. I say they represent the contribution that you make to science projects, and are damn fun to watch as your team propels upwards in the BOINC top 100.


And the downward spiral of rejecting science and becoming a credit whore begins...

But, from a scientific point of view, there appears to be a correlation between the speed and downward force of the spiral and the number of gpus.
26) Message boards : Number crunching : Is anyone glad with the project? (Message 15301)
Posted 14 Mar 2009 by Profile Slicker [TopGun]
Post:
hmmm... death by spam.

Yea, I go out for the day and look what happens:-(

Seems like every thread gets overtaken by Ice, banditwolf, and Misfit in the end, doesn't it?


Yep. Sure does.
27) Message boards : Number crunching : No work (Message 13510)
Posted 1 Mar 2009 by Profile Slicker [TopGun]
Post:
I'm still seeing the "got 0 tasks" message but at least now it is followed by the "reached per CPU-limit of 6 tasks" message. In other words, the longer WUs seemed to have helped.

I'm all in favor of increasing the processing time on the gpu boxes since there is less time spent uploading, downloading, and requesting tasks and more time earning credit. Who knows, we might even get more science done.

And, once they find the real cause of the problem and a complete solution, I'd still be OK with leaving the WUs the same size.
28) Message boards : Number crunching : No work (Message 13344)
Posted 28 Feb 2009 by Profile Slicker [TopGun]
Post:
Milkyway is a no good project! Or, should I say a no good credit project. It is a project. It has good credit. There is no good credit where there is no work. Maybe we should all discuss the credits and the limits some more. What's the point if you can't get any work?

I might as well leave and move to some low credit project like Rosetta until this lack of work issue gets figured out, but that would leave more work for the rest of you. How about if you all leave instead? You'd at least be getting credit then!
29) Message boards : Number crunching : No work (Message 13171)
Posted 27 Feb 2009 by Profile Slicker [TopGun]
Post:
Can I crunch on my scsi card? Its just sitting there idle!!

I've got internal USB 2.0 cards and PS/2 ports doing nothing. What about my power unit? That ought to be cruching with it's own fan an' all...

I've got a sound card that's silent and idle far too often. Give it some work to do! ;P


You are all gluttons for punishment! There aren't enough WUs to keep the cpus and gpus busy and you want to port it to additional hardware? Focus people. Focus!

Every time I checked today, several of my boxes were sitting idle and it takes up to a dozen updates to get them to download work. I don't think the issue is http connections since when they report the completed WUs, that works fine and they still don't get any new work.
30) Message boards : Number crunching : GPU overclocking (Message 12848)
Posted 25 Feb 2009 by Profile Slicker [TopGun]
Post:
I used the CCC "Auto Tune" button. It reported everything passed. Then about 1 in 4 WUs started errored out. I set it back to stock for now. Maybe if I have nothing to do for a whole weekend, I'll slowly increase the gpu speed and keep monitoring closely.

Since the app isn't memory intensive, increasing the memory speed probably won't help much here.
31) Message boards : Number crunching : No work (Message 12712)
Posted 24 Feb 2009 by Profile Slicker [TopGun]
Post:
17K yesterday for one gpu box verses 26.7K on other days. Yep, I'd say it had some problems getting work. I had tried and failed to many times in a row that the scheduler had it set for 3 hours at one point. By forcing a couple manual updates in a row, I finally got it to download again. 30 minutes later, it was back to waiting for 30 minutes.
32) Message boards : Application Code Discussion : GPU app teaser (Message 12128)
Posted 21 Feb 2009 by Profile Slicker [TopGun]
Post:
I've noticed that if I enable both MW and another project, that it will run both but that the MW gpu app will SHARE a cpu with the other project. e.g. Q9450 runs 4 ABC apps and 1 or 2 MW. 3 of the ABC run on their own cpu. The 4th ABC runs on the same cpu as the MW app(s). When the CUDA apps do this, they set their priority to "Below Normal" instead of "Low" (a.k.a. Idle). Any chance the gpu app could be modified to do the same? Then, whatever processing power is left will go to the other app since it will be set to Low.
33) Message boards : Application Code Discussion : GPU app teaser (Message 11903)
Posted 20 Feb 2009 by Profile Slicker [TopGun]
Post:
New install on a Q9450 running XP64 w/ Boinc 6.4.5. It runs 8 gpu WUs at a time. Nice. But, it will only download 12 WUs as if it only recognizes that it has 1 cpu rather than 4 cores. How can I get it to download 12 per core? Or, is it because it is only running on 1 cpu (no other projects) that it will only download 12 at a time?
34) Message boards : Application Code Discussion : GPU app teaser (Message 11902)
Posted 20 Feb 2009 by Profile Slicker [TopGun]
Post:
2. I had to install Microsoft .NET Framework v2.0 for x64. It didn't like that I had .NET v3.5 installed which I thought was supposed to be an extension to 2.0. It wanted 2.0 as well.

I didn't read this line properly.
I already had .net framework 2 installed and thought that was all I needed.
After numerous re-installs of the catalyst drivers & boinc I realised I should also install .net framework 3.5
After it rebooted all is working :)

WOOHOO

running 4 WUs at a time with each taking 13 seconds

Stunning work Cluster Physik


just to recap
I needed both .NET Framework V2.0 and V3.5


Just installed a second card on an XP64 box w/ a Q9450. I ONLY installed .NET 2.0 SP1 and it worked fine. Maybe the issue is that the catalyst driver needs SP1? .NET 3.5 is NOT installed on the box. Hmmm....

Glad it worked for you though.
35) Message boards : Application Code Discussion : GPU app teaser (Message 11821)
Posted 20 Feb 2009 by Profile Slicker [TopGun]
Post:
I think I have a little problem :

20/02/2009 09:21:02 Milkyway@home Sending scheduler request: Requested by user.
20/02/2009 09:21:02 Milkyway@home Not reporting or requesting tasks
20/02/2009 09:21:07 Milkyway@home Scheduler request completed: got 0 new tasks

For the past hour I tried a few times, no workunit for me :(

I will try the 32 bits version, with an overclocking of my HD4850 (and CPU, too ^^)


I upped the limit to 5000, its probably not enough :P Didn't want to jump all the way to 10k and have craziness start happening.


On SETI and GPU grid, after a certain number of errors in a row by the gpu tasks, they stop sending them for a day so as not to cause problems. I blew through about 36 last night really fast before I realized it wasn't working properly. Maybe that's something to add to the todo list as more people try gpu crunching.
36) Message boards : Application Code Discussion : GPU app teaser (Message 11817)
Posted 20 Feb 2009 by Profile Slicker [TopGun]
Post:
I was getting the same error.

I'm also a newbie to ATI crunching and had the same problem last night trying to install on a single core AMD 4000+ running Win2K3 Server x64. I had installed the v8.9 catalyst driver that came with the HD 3870 card. I needed version 8.12. I installed BOTH the control center (48mb) and the wdm (5mb) available here. According to other posts on the board, it sounds like 9.1 doesn't work. You may need to uninstall 9.1 and then install 8.12. After I installed v8.12, it worked like a charm.

Other helpful hints:
1. The ATI WinXP x64 v8.12 driver works fine on Server 2003 x64.

2. I had to install Microsoft .NET Framework v2.0 for x64. It didn't like that I had .NET v3.5 installed which I thought was supposed to be an extension to 2.0. It wanted 2.0 as well.

3. I needed to reboot again after installing wdm even though it didn't ask me to.

Thanks to all those who have worked on the ATI app and for hosting the app on zslip.
37) Message boards : Number crunching : ps_s20_X and ps_s21_X (Message 11182)
Posted 17 Feb 2009 by Profile Slicker [TopGun]
Post:
1. Why does everyone keep comparing credit with the opt app? THAT'S APPLES TO ORANGES!!!! The credit needs to be based on the STOCK MW App to the STOCK SETI app. Just because a majority of MW crunchers are smart enough to install the optimized app and the SETI lemmings are not, doesn't mean MW should be penalized.

2. Why are there andy additional adjustments if the credit is calculated per the flops being done? Is it based on flops or isn't it?

3. If you want to get the credit back in line, my suggestion would be to lower the credit from 12 to 1 for anyone on the LAF team. Just kidding!
38) Message boards : Number crunching : 3rd.in - optimized apps (Message 10670)
Posted 14 Feb 2009 by Profile Slicker [TopGun]
Post:

try using this link to get to Cosmo.

http://www.cosmologyathome.org/testsite/

Much easier on the eyes IMO.


MUCH better


Yeah, I never did understand why they changed it in the first place.

Something to do with Concept of the week.


MW should do App of the week

Travis already has App of the day


yeah. we kinda suck here. so i don't really want to advertise too much considering our record of keeping the server running and work flowing ;P

I wouldn't worry about it. For an alpha project you sure do have a lot of crunchers! If people don't like what has been happening they don't have to do work for the project. Simple really. All up MW has been working fairly well. That is the beauty of BOINC. If one project has problems, just crunch a different for a little while.

ps. Thanks Labbie for the sig ;)


Yep. It is amazing how much crap people will put up with so long as the credit is good.

So, keep the opt apps coming. Keep the credit high enough and people WILL put up with the outages, shenanigans like running test apps in the production environment, speed limits, etc.
39) Message boards : Number crunching : nm_s82_r7/r8 computation errors (Message 10539)
Posted 13 Feb 2009 by Profile Slicker [TopGun]
Post:
If the problem is with the stars files, then how come I can take the stars files from a machine not reporting any errors and copy them to another machine that has about 50% errors and still have the same problem?
40) Message boards : Number crunching : We're back (Message 10118)
Posted 8 Feb 2009 by Profile Slicker [TopGun]
Post:
Work is now downloading from the servers!

Thanks for all the hard work Travis and Dave. We appreciate the tremendous efforts, and can now return good results to you.


Don't be too hasty. I'm pretty sure it'll break in a few minutes :P


Well, it is still working!


Previous 20 · Next 20

©2024 Astroinformatics Group