| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: The Clean Energy Project - Phase 2 Forum
Thread: Research Log: Updates from the Harvard Team |
| No member browsing this thread |
|
Thread Status: Active Thread Type: Sticky Thread Total posts in this thread: 1840
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
A few scientists from Samsung were in town recently for the MRS meeting and they heard about CEP and our research on organic electronic materials. We invited them over and gave them a ‘tour’ of the project and they seemed to like it. We are putting CEP on the map
 .http://www.samsung.com/global/business/semiconductor/ |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Besides the Sony Vaio, a Samsung Galaxy Team... seems to roll very well of the tongue. :D
--//-- |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Dear macusen,
WCG collects about 400y/d in computing time and it has 7 fully active projects and two more on reduced load. That means that an equal share for CEP2 would be about 50y/d. We recently averaged between 14-16y/d and our best daily result ever was at 20y/d. So there is a lot of room for improvement. Dear ibsteve2u and Ascholten, nice idea - a crunching radiator ;). New supercomputing centers actually use this and produce hot water as a side product for the facility heating. Best wishes Your Harvard CEP team |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
An interesting question came up in the forum on the CEP2 share in the WCG computing time. Here is the math: WCG collects about 400y/d and it has 7 fully active projects and 2 more on reduced load. That means that an equal share for CEP2 would be about 50y/d. We recently averaged between 14-16y/d and our best daily result ever was 20y/d. So there is a lot of room for improvement.
https://secure.worldcommunitygrid.org/forums/...30293_lastpage,yes#354402 |
||
|
|
nasher
Veteran Cruncher USA Joined: Dec 2, 2005 Post Count: 1423 Status: Offline Project Badges: 
|
Yes it would be nice if CEP-2 was able to get its mathematical fair share but we know being a manual opt in and with having problems on some multi-CPU machines if more than 1 work unit for CEP-2 is running this will never happen unfortunately.
----------------------------------------for me if I manually took charge of every one of my 3 computer (5 CPU's) instead of only being able to average about 1.5 days per day for CEP-2 i could probably force it up to closer to 3 or maybe a little higher. I do not have that much time on my hands. another way is if BOINC was slightly smarter. the way WCG has to make sure BOINC is only running the max of how many CEP-2 at a time on your computer is by limiting the amount of CEP-2 work units that are allowed on your machine at a time. so instead of getting CEP-2 running on 1 full CPU per computer it only averages about half of that. so the record is 20y/d and we are currently averaging 14-16 y/d. and 50y/d would be about equal share. looking at the project statistics over the time CEP-2 has been running the total average is about 12 1/2 y/d and yesterday was just over 16 y/d. if there was a way to teach BOINC to use the cpu restriction for CEP-2 instead of WCG manually forcing this we would be a closer to the 50y/d but I am guessing we will never get there. since it is a manual opt in most people who are just here at WCG to crunch will not pick up a CEP-2 cause they have not selected it manually. another reason for a little bit of the slump is the 2 new projects prior to them the equal share would have been probably around 60-70 y/d  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
another thing one could try, which I do for one computer is set a profile for cep2 ONLY, put that computer on that profile, tell boinc/wcg unlimited tasks for the computer in the preferences, set the cache to say 7 or 8 days, and load up on CEP2 tasks. You can then turn down your preference to 1 or 2 days then if you want or leave it alone.
This way you have your cache loaded up with CEP2 tasks, only really have to check the computer about once a week to reload it, and it shouldn't run out of tasks. As for CPU utilization, again some will have to play with settings. Aaron |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
We had a brief glitch on one of our servers yesterday at midnight but it didn’t seem to have caused too much trouble. What’s worse is that I didn’t get around to posting our weekend cartoon ;). So here it comes slightly belated.
http://xkcd.com/927/ |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Dear nasher and Ascholten,
CEP2 is more demanding than other projects and is hence not for everyone. So we don't really expect an equal share in WCG computing time. Nonetheless, we see room for improvement of the 3-4% share we have right now. Thank you for your tweaking tips - they may be helpful for our enthusiasts. The mass of casual crunchers, however, probably does not want to be bothered with tweaking, and that's fair enough. A smarter BOINC would be great but we have to work with what we have. The official average of 12.5y/d has to be taken with a grain of salt since it includes the time when we were running Linux only - check out the graphs on the bottom: http://cleanenergy.harvard.edu/index.php?ira=...amp;tipoContenido=sidebar Best wishes Your Harvard CEP team |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
We have passed the mark of 8 million results and also had a great daily return of over 17y/d yesterday. Awesome job everybody - keep crunching!
https://secure.worldcommunitygrid.org/stat/viewProject.do?projectShortName=cep2 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
... and over 30,000 views of this Thread!
 |
||
|
|
|