Welcome to MilkyWay@home

Posts by [boinc.at] Nowi

21) Message boards : Number crunching : Credit drop since starting of the new batch of WU (Message 55928)
Posted 23 Oct 2012 by [boinc.at] Nowi
Post:
The run time of 75 s is a result of running two WU simultaneously on one GPU via app_info.

It is not right to divide the time by two, but, if you want, the run time for one unit is 37.5 s.

This is nearly similar to your run time.
22) Message boards : Number crunching : Credit drop since starting of the new batch of WU (Message 55919)
Posted 23 Oct 2012 by [boinc.at] Nowi
Post:
Hello,

I have mentioned a dramatical credit drop beginning with the start of the new 22_3S-WU. I have observed the running times of the WU on my system (both GPU running two WU simultaneously):

7950:
22_3S 75 s/WU 106,99
09_2s 90 s/WU 159,86

runtime-comparison: 0,83333
credit-comparison: 0,66927
calculated credit on runtime-comparison: 133,22

6950:
22_3S 90 s/WU 106,99
09_2s 130 s/WU 159,86

runtime-comparison: 0,69231
credit-comparison: 0,66927
calculated credit on runtime-comparison: 110,67

So you can see that the credit output of both is lower than expected from the runtimes. Especially for the 7950 with the new GNC-architecture.

The credit drop is forced also by the massive errors per day.

Maybe the devs can look into this.

Cheers

Nowi
23) Message boards : News : New Run Beginning (Message 55656)
Posted 5 Oct 2012 by [boinc.at] Nowi
Post:
These are the WU which are out in the field and send back after error. They will be resent until the maximum number of errors, I mean five, is reached.
24) Message boards : News : New Run Beginning (Message 55652)
Posted 4 Oct 2012 by [boinc.at] Nowi
Post:
I second that! I have the same errors.
25) Message boards : Number crunching : Periodical drop of credits at Milkyway (Message 55201)
Posted 22 Jul 2012 by [boinc.at] Nowi
Post:
Hello,

I observed a periodical drop with my credit output on milkyway. Every five days I encounter a drop of about 200000 credits. You can see it in the following graphic from Free-DC:
https://www.sugarsync.com/pf/D7306442_66157210_921598

If it pleased the server, milkyway was running 24 hours on every day (with and without credit drop). So there was no significant difference in the number of crunched WU, but about 25 % loss on credits.

Has anybody out there encountered similar behavior?
Can anybody explain what happened to me?

Thanks for your replies!

Nowi
26) Message boards : Number crunching : 7850/7870 (Message 54702)
Posted 7 Jun 2012 by [boinc.at] Nowi
Post:
You can use this app_info.xml to run two tasks parallel. Kick them out of phase and you will have lesser utilization-drops.

<app_info>
<app>
<name>milkyway</name>
</app>
<file_info>
  <name>milkyway_separation_1.02_windows_x86_64__opencl_amd_ati.exe</name>
  <executable/>
</file_info>
<app_version>
  <app_name>milkyway</app_name>
  <version_num>102</version_num>
    <flops>1.0e13</flops>
    <avg_ncpus>0.05</avg_ncpus>
    <max_ncpus>1</max_ncpus>
    <plan_class>opencl_amd</plan_class>
    <coproc>
      <type>ATI</type>
      <count>0.5</count>
    </coproc>
    <cmdline></cmdline>
  <file_ref>
   <file_name>milkyway_separation_1.02_windows_x86_64__opencl_amd_ati.exe</file_name>
   <main_program/>
  </file_ref>
</app_version>
</app_info>


Cheers

Nowi
27) Message boards : Number crunching : GPU-task not ending (Message 54237)
Posted 29 Apr 2012 by [boinc.at] Nowi
Post:
I have tried 12.3 and 12.4 but it was no solution of the problem.

I posted this in the AMD and OpenCL-dev forum:

Hello,

I have problems with every newer driver than the release driver (8.921.2-120119a-132101E-ATI). I encounter the following problems with the OpenCL-Application:

Some tasks are not ending or send to CPU instead.
Often I encounter the message: "amdkmdap not responding and was restartet" (translated from german)
The screen is gone black and cannot be restored except of doing a reset.

System Information:

C2Q 9550 @ 3.4 GHz, Asus P5Q-E, 8GB, Win7 64 Bit, Boinc-Version 7.0.25

2 Gigabyte 7950 WF3 (not crossfired when using BOINC, second card has a dummy plug instead of a display)


Hope that I will get an answer.

Do you have answers here, too.
28) Message boards : Number crunching : MilkyWay tasks always on high priority (Message 54191)
Posted 27 Apr 2012 by [boinc.at] Nowi
Post:
Thanks for all your answers!

In fact running of M@H-tasks in high priority has become more and more an "optical" problem for me. Roll back to version 6 isn´t really an option, because I´m running long lasting CPDN tasks, which will be lost then.

I will wait for the fix in 7.0.27.
29) Message boards : Number crunching : MilkyWay tasks always on high priority (Message 54140)
Posted 25 Apr 2012 by [boinc.at] Nowi
Post:
Hi John,

thank you for this information!

Anyone who wants can find new versions here: http://boinc.berkeley.edu/dl/?C=M;O=D
30) Message boards : Number crunching : MilkyWay tasks always on high priority (Message 54104)
Posted 22 Apr 2012 by [boinc.at] Nowi
Post:
Hello,

on friday boincmanager showed me an outage on the milkyway-server (project down for maintenance). So my backup-projects starts working (first donate, than distrrgen). After about one hour I received new WU from milkyway, but from then all tasks running in high priority, although milkyway is the only gpu-project currently running (by now distrrgen is backup-project with resource share 0).

This high priority issue remains since more than two days.

BM 7.0.25 (Win 7)

Do you have any suggestions for me?

Thanks

Nowi
31) Message boards : Number crunching : GPU-task not ending (Message 54058)
Posted 18 Apr 2012 by [boinc.at] Nowi
Post:
Thank you! I will give the driver a try.
32) Message boards : Number crunching : GPU-task not ending (Message 54052)
Posted 17 Apr 2012 by [boinc.at] Nowi
Post:
Hello,

from time to time I got very long and, if not cancelled before, longer than 100 % running WU on my 7950 GPU. On example is:

http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=178798443 (aborted by me today in the morning (Austian time)) Please look in the stderr.out :)

It seemed that the WU ran on the CPU instead of the GPU.

I´m using the 1.02 OpenCL-client with anonymous platform for running 2 tasks parallel. Driver is the release driver of 7950: 8.921.2-120119a-132101E-ATI
Boinc 7.0.25
Win 7
Two 7950 not crossfired for BOINC.

Do you have any suggestions?

Thanks

Nowi
33) Message boards : Number crunching : ATI HD 5870 not being recognized (Message 53684)
Posted 15 Mar 2012 by [boinc.at] Nowi
Post:
Hello,

as I understand you have a notebook with two 5870. None of the AMD-mobile graphic cards supports double precision. The naming as 5870 is IMHO irritating.

Sorry!
34) Message boards : Number crunching : Multi-GPU Question (Message 53542)
Posted 4 Mar 2012 by [boinc.at] Nowi
Post:
For a first information look here: http://en.wikipedia.org/wiki/AMD_CrossFireX
35) Message boards : Number crunching : Updates not happening automatically (Message 53416)
Posted 25 Feb 2012 by [boinc.at] Nowi
Post:
I´m using 7.0.18 with a pretty full cache for both, GPU (M@W) and CPU (several projects), by using this network preferences:

Minimum work buffer: 1.00 days
Max. additional work buffer: 0.01 days.

It runs pretty smooth. Try this... :)

Cheers

Nowi
36) Message boards : News : Separation updated to 1.00 (Message 53040)
Posted 10 Feb 2012 by [boinc.at] Nowi
Post:
You can find the version here: http://boinc.berkeley.edu/dl/?C=M;O=D
37) Message boards : News : Separation updated to 1.00 (Message 53037)
Posted 10 Feb 2012 by [boinc.at] Nowi
Post:
No, it is really high cpu-usage according to the windows task-manager which shows usage up to 25 % on my quad-core for each milkyway process. Boinc shows 0.05 as expected. Please show also in my example wu in the original post. The cpu-time is very high!
I know the AMD drivers had the high CPU problem in 11.7 and 11.8 but I guess that could be different on different hardware. Does it happen with the current 12.1 drivers for you?

I tried 11.11c, 12.1, 12.2 preview. Every driver showed the high cpu-usage.
38) Message boards : News : Separation updated to 1.00 (Message 53027)
Posted 10 Feb 2012 by [boinc.at] Nowi
Post:
No, it is really high cpu-usage according to the windows task-manager which shows usage up to 25 % on my quad-core for each milkyway process. Boinc shows 0.05 as expected. Please show also in my example wu in the original post. The cpu-time is very high!
39) Message boards : News : Separation updated to 1.00 (Message 53017)
Posted 10 Feb 2012 by [boinc.at] Nowi
Post:
Hello,

Opposite to my first post above, I have encountered problems now. I have very high cpu-usage with 1.00 and 1.02 app and two 5850 GPU.

I faced the problem after upgrading an old ATI-beta-driver from December, which worked well for playing Skyrim, but shows intermittent BOD. With this driver Milkyway runs with very low cpu-usage very fine. I deinstalled the beta driver, cleaned with driver-sweeper as I always do when installing new ATI-driver, reboot, install new driver (first attempt 12.2 preview), reboot, starting boinc. Result: Milkyway is starting with high cpu-usage (M@H wants one core).

I tried several drivers installed in the above scheme. 11.11c, 12.1, reverting to old (previous functioning) beta-driver, 12.2 preview. Tried app_info w/o cmdline and without app_info. Changed crossfire enabled/disabled. Still the same result M@H wants a whole core.

example wu: http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=100185622

Collatz and Moo are running fine with every driver.

Boinc 6.12.34 (upgrading to 7.0.x shows no effect)
Win 7 64 Bit

Do you need further information?

I hope for suggestions from you.

Thanks!

Nowi
40) Message boards : News : Separation updated to 1.00 (Message 52932)
Posted 9 Feb 2012 by [boinc.at] Nowi
Post:
New version runs fine for me! Run-times about 8 s faster than 0.82 separation! Thanks Matt!

System
Win 7 64 Bit
6.12.34
2x 5850
app_info with two WU per GPU


Previous 20 · Next 20

©2024 Astroinformatics Group