Posts by zombie67 [MM]
log in
1) Message boards : Application Code Discussion : Intel Open CL (Message 60171)
Posted 19 Oct 2013 by zombie67 [MM]
No, Intel HD 4000 does not support DP.
2) Message boards : News : post scheduling issues here (Message 59576)
Posted 11 Aug 2013 by zombie67 [MM]
A workunit is flagged with "4 CPUs" and will not start unless I set preferences to use all 4 processors on my system.

I normally allow 2 processors (out of 4) to work on SETI and Milkyway projects during the day and allow 4 processors when I leave the office.

Right now nothing is running during the day because of this "4 CPU" setting, I am assuming, and no other work from either project has been scheduled for me because of the size of this workunit for Milkyway - 700+ hours.


Tasks are assigned the number of CPUs at the time of the download. So you had all 4 CPUs available when those MW tasks were downloaded. No way to change that to 2, after the tasks have been downloaded, aside from deleting them and downloading fresh tasks with your settings to allow only 50% CPU. After downloading, you can change your settings to use 100%. The MT tasks will still use only 2 cores. Of course, the scheduling logic may cause BOINC to run two MW tasks, or 4 SETI tasks, or 1 MW task and two SETI tasks.
3) Message boards : News : post scheduling issues here (Message 58827)
Posted 13 Jun 2013 by zombie67 [MM]
Just curious: Why both single-thread and multi-thread apps? I am getting sent both, and that leads to idle threads. For example, on an 8 thread machine, I get sent 5 ST tasks and 5 MT tasks. When the ST tasks run, it leaves 3 threads idle. And BOINC won't ask for more ST threads because it thinks it already has enough work due to the queue of MT tasks.

If both apps will stay, how about a way to select one or the other on our preference page, like we can do with GPU types?
4) Message boards : Number crunching : N-Body CPU: does it checkpoint? (Message 57468)
Posted 10 Mar 2013 by zombie67 [MM]
I have one running, and I need to know if a power cycle will kill it.
5) Message boards : Number crunching : Not getting WU's (Message 52509)
Posted 14 Jan 2012 by zombie67 [MM]
I am not running DHCP so ipconfig /release and ipconfig /renew have no meaning for my machine.


A detatch (not reset) and reattatch will do it.

Regards
Zy


Yeah, that worked for me.
6) Message boards : News : ATI 58x0 GPU fix released (Message 38269)
Posted 7 Apr 2010 by zombie67 [MM]
Can't get work to try it out. "No work available"

Edit #1: Required a Project Reset. Odd. A newer app version should not require manual resets like that. Anyway, I have new work now.

Edit #2: seems to be working. Thumbs up!
7) Message boards : News : ATI 58x0 GPU fix released (Message 38253)
Posted 7 Apr 2010 by zombie67 [MM]
When will it be available as a stock app? I don't use app_info.xml.
8) Message boards : News : testing new validator (Message 38046)
Posted 5 Apr 2010 by zombie67 [MM]
Just counting pages at 20 tasks each, I am currently at 9 valid and 2 invalid. 82% valid tasks, is not going to work in the long run, obviously. But I'll hang around for the shakedown.
9) Message boards : News : testing new validator (Message 38035)
Posted 5 Apr 2010 by zombie67 [MM]
Seems to me that someone should do a standalone test and compare results between 48xx and 58xx again, just to make sure everything works properly.


And if they do produce the same results, then perhaps the validator needs to be tested as well.
10) Message boards : News : testing new validator (Message 38033)
Posted 5 Apr 2010 by zombie67 [MM]
Ouch. If correct, that's going to make this messy. Does HR (homogeneous redundancy) exist for GPUs?

Edit: From what I have just learned, the answer is maybe.
11) Message boards : News : testing new validator (Message 38030)
Posted 5 Apr 2010 by zombie67 [MM]
Okay, I've run across a few with:

minimum quorum 3
initial replication 5
max # of error/total/success tasks 1, 6, 6

http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=90198132
http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=90151888

Max # of error 1....is that really what you want? Also, why is my machine marked as invalid? It is using the stock app, with a 5870.
12) Message boards : News : testing new validator (Message 38027)
Posted 5 Apr 2010 by zombie67 [MM]
Most of mine (edit: many, not most) are ending up with "Completed, validation inconclusive". 4 results, with a max of 4, and the no one gets any credits. Something doesn't seem right with this scheme.

http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=82607813

http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=87913982
13) Message boards : News : testing new validator (Message 38022)
Posted 5 Apr 2010 by zombie67 [MM]
Validation will now work as follows: Every result that could improve one of our searches will be validated (with a min quorum of 3[...]


Why a quorum of 3? Why not two, and if they don't match closely enough, a third task is sent? It seems like a waste of resources.
14) Message boards : Number crunching : Suggestion for downtime (Message 36382)
Posted 10 Feb 2010 by zombie67 [MM]
MW only has 1 server.


I am not talking about the hardware. I am talking about the various BOINC software apps. Sorry if I was not clear. You can see them here:

http://milkyway.cs.rpi.edu/milkyway/server_status.php


So in this case, the hardware is obviously not down, or we would not be posting. And we would have gotten an RSS post saying that the "feeder" is down (among others).
15) Message boards : Number crunching : Suggestion for downtime (Message 36369)
Posted 10 Feb 2010 by zombie67 [MM]
Whenever the one of the various MW BOINC servers (upload/download, scheduler, etc) goes down, have a deamon automatically submit an RSS post. That way our RSS readers would pick it up right away, and we can take appropriate action with our machines (move them temporarily to other projects).

You could even use such a thing to send a text message to the project admin.

Of course, this would not be necessary if it would be allowed that the queue be more than 30 minutes worth of work for fast GPUs.
16) Message boards : Number crunching : Another slooow down in the works?? (Message 35873)
Posted 21 Jan 2010 by zombie67 [MM]
Bigger tasks for GPUs. It's free comared to upgrading hardware.
17) Message boards : Number crunching : Now that we have native ATI GPU support, how about longer tasks? (Message 35751)
Posted 17 Jan 2010 by zombie67 [MM]

In my opinion, you should consider looking at whether or not there is a way to use Homogeneous Redundancy classes to separate GPU from CPU and give GPUs the longer task and leave the shorter tasks to CPUs.


HR isn't a factor here. With a quorum of 1, you don't use multiple replications for validation.
18) Message boards : Number crunching : Now that we have native ATI GPU support, how about longer tasks? (Message 35703)
Posted 16 Jan 2010 by zombie67 [MM]
Thanks for implementing native ATI support!

And now that we have it, how about issuing tasks exclusively for ATI GPUs that run (say) for an hour (on a 4870)?

No change in credits/hour. That way we can fill up a normal queue of work lasting (say) a day or two.

In addition to letting us to weather downtime or network issues, it would DRAMATICALLY drop the load on the project server and network load.
19) Message boards : Number crunching : Testing ATI Application Availability (Message 35457)
Posted 11 Jan 2010 by zombie67 [MM]
A quick question, why does the app report that it uses 41% of the CPU?


Mine is showing 1.02 CPUs. Ouch.

Meh. It should be a purely cosmetic misrepresentation that can be corrected later. Are you not currently getting full use of all your CPU cores for other BOINC tasks?


It *should* be?
20) Message boards : Number crunching : Testing ATI Application Availability (Message 35454)
Posted 11 Jan 2010 by zombie67 [MM]
A quick question, why does the app report that it uses 41% of the CPU?


Mine is showing 1.02 CPUs. Ouch.


Next 20

Main page · Your account · Message boards


Copyright © 2017 AstroInformatics Group