| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: Suggestions / Feedback
Thread: Smaller work units for older systems |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 3
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I am using two systems: a fairly new one with an Intel Core Duo 2.4GHz and an older on with an AMD Athlon .
----------------------------------------The older one receives the same size of work units as the newer one, which leads to not meeting the deadline for the delivery of the work unit because this PC doesn't run 8 hours a day (in fact I switch it on 2-3 times a week for about one hour). Working on units I won't get credit for (and which will be distributed to others again) doesn't make sense so I am going to stop WCG on that system. If I could select to receive smaller units which can be completed in about 1 hour instead of 12-15 hours that system could participate and provide value for WCG. Even if it's a small one, I assume there are more people like me and it would finally add up to something that makes a difference. Selecting the size (=time needed) of work units should be configurable per system, because for one system it makes sense to receive larger amounts if is switched on for a longer time and for others it makes sense to receive smaller amounts. [Edit 1 times, last edit by Former Member at Feb 14, 2009 12:14:35 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
You should try making a new profile for the older machine and select only the Nutritious Rice project. They run at 10 hours regardless of the system. It gets as much done as it can and then reports after 10 hours.
And actually the techs announced recently that the jobs were going to be 6 hours crunch time. So that project would probably be your best bet. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
A look into the future discussed in other threads [test your search skills for posts by knreed] is to develop an function for the work generators & feeders to dynamically and automatically cut and assign "work according device crunching power". A supplemental option to go with that, if feasible, is to provide a device profile setting to express the preference to receive in augmentation to the automatic function shorter, normal, longer work units... and no, not "per project". There's a limit to what's practical, no limit to our dreams.
----------------------------------------Both these functions require BOINC server software development wish has only been done in outline, so it is still a very long way away... 2010, maybe. Generally, all but CEP jobs are now set to run on a project average of 6 hours, with the exception of RICE where the newest batches have a fixed 6 hour run time irrespective the power of the device. If you look in the support FAQ's, see link below, there's a project resource matrix, but monitoring the project graphs, see link below, provides a more up to date view of how job durations are currently averaging. cheers
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
|