Welcome to MilkyWay@home

Posts by Waffleironhead

1) Message boards : News : Nbody WU Flush (Message 72806)
Posted 15 Apr 2022 by Waffleironhead
Post:
I am now consistently getting separation WU's.
Thank you!
2) Message boards : Number crunching : How to not Get Multiple CPU Units (Message 71869)
Posted 4 Mar 2022 by Waffleironhead
Post:
AFAIK N-body is always going to be multi-threaded.
To stop getting them you need to go into your account and click on: "Preferences for this project MilkyWay@home preferences"
Then uncheck "n-body"
3) Message boards : Number crunching : New Benchmark Thread - times wanted for any hardware, CPU or GPU, old or new! (Message 69639)
Posted 1 Apr 2020 by Waffleironhead
Post:
Another point of data for you. got to play around with a tesla p100

NVIDIA Tesla P100-PCIE-16GB

average for a couple thousand WU's is 11 seconds per unit.
4) Message boards : Number crunching : old 32 bit and hd6950 (Message 66222)
Posted 2 Mar 2017 by Waffleironhead
Post:
The 6950 will run just fine.
To get the most out of it though you should consider running multiple WU at once. I have found on my 6950 that running 8 wu at once provides the best results.
There is a thread in this same forum about running multiple WU at once.
My app config looks like this:

<app_config>
<app>
<name>milkyway</name>
<gpu_versions>
<gpu_usage>0.125</gpu_usage>
<cpu_usage>0.125</cpu_usage>
</gpu_versions>
</app>
</app_config>
5) Message boards : News : MilkyWay@home Version 1.38 Released (Message 65231)
Posted 21 Sep 2016 by Waffleironhead
Post:
I'm noticing quite a few more invalids from "completed can't validate" since the new version came out. Over 100 right now.

Its hard to tell if there is a problem with the program or my system when the WU are not given a chance to validate.

Would it be too much to up the replication to 10 or more?
6) Message boards : Number crunching : "Completed, can't validate" - what's this? (Message 64997)
Posted 6 Aug 2016 by Waffleironhead
Post:
Other projects have a higher threshold, Maybe the admins can set the limit to 10? 25?
7) Message boards : Number crunching : threads per gpu? (Message 64963)
Posted 31 Jul 2016 by Waffleironhead
Post:
I'm running 16 units at the same time on my 6950.
8) Message boards : Number crunching : AMD GPU setup (Message 64773)
Posted 29 Jun 2016 by Waffleironhead
Post:
I have been running 2 units at a time on my hd6950.
I use an app_config.xml file in the folder.
\ProgramData\BOINC\projects\milkyway.cs.rpi.edu_milkyway


Contents of file for the MilkyWay@Home application is:

<app_config>
<app>
<name>milkyway</name>
<gpu_versions>
<gpu_usage>0.5</gpu_usage>
<cpu_usage>0.5</cpu_usage>
</gpu_versions>
</app>
</app_config>
9) Message boards : Number crunching : Why is all credit the same? (Message 64231)
Posted 7 Jan 2016 by Waffleironhead
Post:
The points being awarded are based on the amount of work being done, not the amount of time it takes. Since these units you are referring to are the same size computation wise(same number of flops) they are all awarded the same points.
Other projects have varying size work units so have to employ a varying point system.




©2023 Astroinformatics Group