Posted 5 Aug 2010 by
Paul D. Buck
hi,
I use boinc 6.10.58 on Win XP. Yesterday I got an new wu and it were 19% complete. I turned my pc off and went to sleep.
Now I started my pc and boinc today. The new were set up to 100% complete and I got credit and the next wu.
The old Units ever started again at the point I stopped it.
Is this an new bug?
Michael
This is normal.
I made a modification in the new applications. The workunits will finish after they run 12 hours or 13 hours after the start of computing (even if the progress is less than 100%).
Credits are granted depending of the run time.
Ok, I will keep watch on the tasks to see if they over-run 25 hours ... the last one I aborted had not exceeded that threshold ... thanks for the tip.
Posted 4 Aug 2010 by
Paul D. Buck
Ver 1.35 on Win7-64 the new tasks seem to want to run for 90+ hours... I don't think this is correct. Pls advise...
BTW, my Mac's task seems ready to complete in about 12 hours as should be normal.
Posted 28 May 2010 by
Paul D. Buck
I have a WU that has been running for 92 hours. I am concerned that if I stop it now, I have wasted all those hours. As it is, 92 hours equates to $15.00 of Electricity cost for me. If I only get 50 points for this, Is it worth crunching?
Yes ...
It is "free" in that the load on your CPU is nil ... so, though it "runs" for 12 hours, it uses only seconds of CPU in that entire period ...
Over time I hope the data collection and analysis improves to the point where some real utility comes out of this project ... but, gathering the data is key ...
Posted 10 May 2010 by
Paul D. Buck
Nothing on CPU Boinc
I suspect that's the reason, I don't think this project looks at GPU work units yet.
It must, I only run Collatz, MW, GPU Grid and DNETC on my GPUs (5 systems, including OS-X), and it has recorded data on these projects ... of course, it does not distinguish these tasks as being run on the GPU, it records them, apparently against the CPU for purposes of recording. So, there will be data skew in the sense that some CPU tasks for Collatz (for example) on OS-X will be recorded as running in `58 minutes and yet the same class task on the same class CPU will also be recorded as taking several hours ... depending on if the participant is running them on the CPU side or on the GPU (as I am)...
Likewise, when I was still running MW on my CPUs they were taking I think up to 4 hours to run on an i7 920 ... yet, same i7 I have is going to record them at 2 minutes on one system and 5 minutes on another, same class CPU, yet the data will be spread from minutes to hours all because the tasks can be run on the CPU, or in my case on HD5870, HD4870, GTX295, or GTX280 ... so, in that sense the data is there, but, the data is not properly collated and can give misleading information ...
On "pure" CPU only projects that are running on a class of CPU the data can be trusted with the usual caveats ...
Then again, over time, they may figure out how to capture enough data to start to segregate the data by GPU as well ... of course there the issue will be that there will still be a spread because of the number of people that have factory OC cards and those that push those limits even higher ...
Posted 2 May 2010 by
Paul D. Buck
How do you set to run 25 workunits concurrently?
FreeHAL allows you to make a preference setting that allows up to 25 tasks to be downloaded and run at the same time... it exploits a side effect in the way UCB does NCI projects running on the PC ... so, you can run 25 FreeHAL NCI tasks, the one here, and Anansi (if and when they start issuing tasks again), and QCN all at the same time you are running n tasks for other projects (where n=the number of CPUs you have plus the number of GPUs)... on W02 for example I do just that ... so I have 25 FreeHal, 8 CPU tasks, QCN, and WUProj and 2 GPU tasks all running at the same time for a total of 37 tasks ... on other systems I only have 36 because I don't have QCN running on them ...
If more NCI projects come along you can add them as well ... you will lose a little on the side running these "extras", but the cost is minimal ... I expect that my migration to 64 bit OS about the same time as adding some of these extras will more than make up (64-bit applications can be faster for some projects like ABC).
Posted 18 Apr 2010 by
Paul D. Buck
Thank you... looks nice ... I have my first two tasks done as well ... with more on the way!
Thanks ...
Posted 17 Apr 2010 by
Paul D. Buck
I have had to turn off Norton's "Sonar" because it seems to be flagging the application on Vista and Win XP though it may be that I need to update to the latest suite ...
Just a note to advise ... one system I hit the daily limit already ... not a problem as it will cure itself soon enough ... :)
In the mean time I will take a dive into seeing if it is because of Norton version ...
Posted 17 Apr 2010 by
Paul D. Buck
Most of the projects that use only part of the CPU show in the status column a faction of CPU use which is mostly (for me anyway) just a reminder that this is an Non-CPU Intense task... in other words it is a nice thing to do ...
For example, I just started running the app on two systems until it looks like it will work and then I will add my other two systems to the mix ... and on the Win7 machine, for example I have the nice "flags" for the ATI task and the Nvidia task and the Quake Catcher Network task ... but not the WUProp task ... (well FreeHAL is also missing the flag, but, that is not a problem for here)...
Anyway, in your copious free time ... t'would be a nice touch to add the "(0.01 CPU)" status marker ... if'n you have no clue I can mouse about and see if I can find the reference for the flag ...
It may only require a change to the "plan class" see: AppPlan
Also see: Non-Cpu Intensive
INteresting, the first time I posted this the anti-spam flagged it as spam ... I pulled out the two links and it was fine with it ... edited it to see which link it did not like and this is the 3rd edit and it is ok with it now ... odd ...