| 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: short and sweet |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 11
|
|
| Author |
|
|
kateiacy
Veteran Cruncher USA Joined: Jan 23, 2010 Post Count: 1027 Status: Offline Project Badges: 
|
E201405_ 999_ C.14.C11H8N2S.00000037.3.set1d06_ 1-- 640 Valid 3/5/11 10:53:14 3/5/11 23:25:07 1.69 40.7 / 41.5 E201405_ 999_ C.14.C11H8N2S.00000037.3.set1d06_ 0-- 640 Valid 3/5/11 10:37:02 3/5/11 21:04:03 1.14 42.2 / 41.5 This is the quickest CEP2 WU I've ever seen -- both my wingman and I completed all 16 jobs, in 1.14 and 1.69 hrs.  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Dear kateiacy,
This actually has an easy explanation - we have now started sending out a batch of smaller molecules again. For the next few days there will be quite a few fast wus. A rough indication of the size of a molecules/time the wu takes is this part of the ID string. E201405_ 999_ C.14.C11H8N2S.00000037.3.set1d06 14 is very small, the last molecules of the 'A' series were between 28 and 31. But don't be fooled: The computing time does not grow linearly but to the 3rd or 4th power, i.e., if the molecule size doubles, the time will increase by a factor of 8 to 16. So enjoy the small work units while they last  .Best wishes, Your Harvard CEP team |
||
|
|
anhhai
Veteran Cruncher Joined: Mar 22, 2005 Post Count: 839 Status: Offline Project Badges: 
|
Yes, I have also noticed several WUs that are finishing much much faster. The WUs are also finishing all 16 jobs.
---------------------------------------- |
||
|
|
kateiacy
Veteran Cruncher USA Joined: Jan 23, 2010 Post Count: 1027 Status: Offline Project Badges:
|
Dear kateiacy, This actually has an easy explanation - we have now started sending out a batch of smaller molecules again. For the next few days there will be quite a few fast wus. A rough indication of the size of a molecules/time the wu takes is this part of the ID string. E201405_ 999_ C.14.C11H8N2S.00000037.3.set1d06 14 is very small, the last molecules of the 'A' series were between 28 and 31. But don't be fooled: The computing time does not grow linearly but to the 3rd or 4th power, i.e., if the molecule size doubles, the time will increase by a factor of 8 to 16. So enjoy the small work units while they last .Thanks for the detailed explanation! We curious crunchers appreciate your responsiveness to our questions! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
[
 ] S&S they may be, like shorties too, but had to tune up the cache to quadruple the number allowed to max 16, to get the right mix or have CEP2 run once every 10 hours for 2 hours run time and a 8 hour gap... the price of variability. The 2 hour jobs are good for 10Mb upload, the 10 hour jobs 31-32Mb i.o.w. same amount of crunching increased bandwidth use to 50Mb... 66%.Happy Crunching notwithstanding ... science is never hurdleless ;>) |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
... the good news is that the relative efficiency in crunching is pretty much the same. The checkpoint 3 disk whirl is much shorter. Every cloud has a lining somewhere.
Forward On. PS, Bandwidth use is pretty much limitless from my world corner. Telecom Italia is good. |
||
|
|
Speedy51
Veteran Cruncher New Zealand Joined: Nov 4, 2005 Post Count: 1326 Status: Offline Project Badges: 
|
E202076_ 690_ C.24.C22H14N2.00168460.3.set1d06_ 0-- Took 3.05 hours on a stock 980X Skipped job #15 (5016): called boinc_finish
---------------------------------------- [Edit 1 times, last edit by Speedy51 at May 10, 2011 5:42:17 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Look at the http://bit.ly/WCGCE1 chart and observe the steady increase of run times since starting the new batch... we're on 7.5 hours mean now.
Crunch On --//-- PS, my Linux box recorded 96.5% today for the 2 concurrent ''overnights''... nearing maximum achievable again, for this device :D |
||
|
|
Speedy51
Veteran Cruncher New Zealand Joined: Nov 4, 2005 Post Count: 1326 Status: Offline Project Badges:
|
I finished E202416_ 576_ C.26.C21H12N4O.00711718.2.set1d06_ 1 in 2.97 hours I returned on the 15th. Finished up to job # 11. Still in pv state. Also returned E202425_ 338_ C.24.C19H9NS2Se2.00465670.0.set1d06_ 1 in 6.33 hours today 16th) finished to job # 15. Project Mean is 7.17 Hours/Task 7.89 as of the 16/6
---------------------------------------- |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Nice, Speedy51, the last S&S me cruncher had was 4.18 hours on June 9. They usually have one of those RC= early retirement log entries.
----------------------------------------Happy Crunching (working on year 2 now :P) --//-- edit: Present project average is 7.8 hours, with a slight trend down in the past few days from 8.0 hours. Means little as that could be just a change in the device contribution mix... see http://bit.ly/WCGCE1 [Edit 1 times, last edit by Former Member at Jun 17, 2011 8:08:23 AM] |
||
|
|
|