Posts by apohawk
log in
1) Message boards : News : Updated Server Daemons and Libraries (Message 65298)
Posted 26 Sep 2016 by Profile apohawk
apohawk,

Was that one of the cards that needed a custom app_config file before????

Also can anyone confirm if they get WUs for the new rx480/470s?

Jake
Yes, it wasn't getting WUs before update. Now it gets WUs out of the box.
2) Message boards : News : Updated Server Daemons and Libraries (Message 65287)
Posted 26 Sep 2016 by Profile apohawk
I'm finally getting WUs for r9 fury, well done!
So far i got 61 validation inconclusive, 5 validated correctly and no errors.

Thank You, Jake
3) Message boards : News : MilkyWay@home Version 1.38 Released (Message 65229)
Posted 21 Sep 2016 by Profile apohawk
I still can't get WUs for f9 fury out of the box. :(

2016-09-21 21:07:58 | Milkyway@Home | Sending scheduler request: To fetch work.
2016-09-21 21:07:58 | Milkyway@Home | Requesting new tasks for AMD/ATI GPU
2016-09-21 21:08:00 | Milkyway@Home | Scheduler request completed: got 0 new tasks
4) Message boards : News : GPU Issues Mega Thread (Message 65121)
Posted 9 Sep 2016 by Profile apohawk
apohawk,

Our application does currently support your GPU, the issue is our server does not recognize it as a GPU that can crunch our work units. You can fix this short term buy using the config files others have posted on number crunching. I am hoping that the release I am working on now will fix the issue, but I can not make any promises. If it doesn't, I will keep looking for a solution.

Thanks. Yes, i crunched few WUs i had in queue before removing 6950 and they were crunched without any changes, app_info, any problems and validated properly.

Any ETA on that release you're working on?
5) Message boards : News : GPU Issues Mega Thread (Message 65116)
Posted 8 Sep 2016 by Profile apohawk
I posted in 'numbers cruncing' forum, but got no response so far, so i will report here as well.
Please, add support for amd r9 fury and other new GPUs.
6) Message boards : Number crunching : Can't get work (R9 Fury Nitro) (Message 65112)
Posted 7 Sep 2016 by Profile apohawk
Hello. Recently (31.08.2016) i bought new GPU, R9 Fury Nitro and put it to work. I updated graphic drivers, crunched some MW WUs from queue and moved on to other projects to see how it works.
Then i tried to do more testing with MW and i couldn't get any WUs. Server status says there is work, but i just get 0 WUs, as logged below:
2016-09-07 21:12:22 | Milkyway@Home | Sending scheduler request: Requested by user.
2016-09-07 21:12:22 | Milkyway@Home | Requesting new tasks for AMD/ATI GPU
2016-09-07 21:12:24 | Milkyway@Home | Scheduler request completed: got 0 new tasks

Host is https://milkyway.cs.rpi.edu/milkyway/show_host_detail.php?hostid=176819.
I don't use anonymous platform (app_info.xml), should i?
Is there any general problem with getting work recently, or with r9 fury?
7) Message boards : Number crunching : nbody (Message 41633)
Posted 21 Aug 2010 by Profile apohawk
When will ATI application for those be available?
8) Message boards : Number crunching : more credits :) (Message 37098)
Posted 9 Mar 2010 by Profile apohawk
Alright, thanks. Hopefully that will render current abort craziness pointless.
9) Message boards : Number crunching : more credits :) (Message 37096)
Posted 9 Mar 2010 by Profile apohawk
Then they will switch to aborting those, which will still burden the project's servers.
What should happen is reduce granted credits of those shorter WUs accordingly to their runtime and some kind of repercussions for those sabotaging projects work by abusing this (hopefully) temporary gap.
Just a clear message to those abusing project's resources, that such behavior won't be tolerated.
10) Message boards : Number crunching : Is it all about the GPU ? (Message 37092)
Posted 9 Mar 2010 by Profile apohawk
It should be the same. CPU part is estimated to be barely 0.05 (initialization and feeding data to the GPU i guess), so just focus on GPU.
11) Message boards : Number crunching : more credits :) (Message 37090)
Posted 9 Mar 2010 by Profile apohawk
It seems to me that the cheating is already going on and due to the massive aborting of longer WUs, some of them get never crunched, wasting project's resources.
http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=70408603
Here is this scenario. WU aborted by one abort-happy user, then by another, results in wasted WU.
Come on, all aborted WUs one minute after receiving them ? Hosts don't display owner (anonymous), so it's up to project admins to search the database.
How much the ratio of such permanently errored WUs increased due to this practice i wonder.
12) Message boards : Number crunching : Is it all about the GPU ? (Message 37088)
Posted 9 Mar 2010 by Profile apohawk
Check in BOINC Manager's preferences "Use GPU while computer is in use" and "Only after computer has been idle for ____ minutes". Status "waiting to run" indicates that the WU was started and stopped, most likely due to the settings above.




Main page · Your account · Message boards


Copyright © 2017 AstroInformatics Group