| 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 |
|
|
Amelior Scout
Cruncher United States Joined: Dec 20, 2011 Post Count: 17 Status: Offline Project Badges: 
|
OOO! I saw something similar to this a while back! It was something to do about kinetic energy collected from people walking on the sidewalks.. Awesome! I've been wanting to hear the developments on this type of tech.. Thanks I'm really glad you posted this! :D
----------------------------------------
~Life is ever-changing~
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Here are some more details to the White House Materials Genome Initiative from the other day.
http://www.whitehouse.gov/sites/default/files...fact_sheet_05_14_2012.pdf |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
All right, we have a bit of a backlog on this thread – so let’s get to it:
Dear Sgt.Joe and 5pot: The computational cost scales roughly with the 3rd or 4th power to with a number of electrons in a molecule and the number of basisfunctions necessary to describe it, respectively. We primarily use the latter in the medium wu basis set. We went up to about 530 functions in a DZP basis before, currently we run molecules with about 480 and we are pretty sure we can go up to 550. 600 should also be possible, but at that point we’ll have to assess how our cruncher machines cope with the stress, i.e., whether I/O and network starts to make trouble and whether we still finish a sufficient fraction of the workunits. Once the molecules get too big for the grid we will change to different workunit design, i.e., we will study different aspects of the candidates. On the cluster we ran up to 1200 bf molecules but there are only a few cases. We mostly crunch on 700-800 bf molecules right now. The cluster machines have 10s of GB of RAM and are designed for heavy tasks. Richard – sorry for the tardiness, we didn’t mean to ignore you. Yes, this is interesting science, but not quite aligned with our expertise. But we always appreciate your posts  Best wishes from Your Harvard CEP team |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
CNET has also picked up on the Materials Genome Initiative!
http://news.cnet.com/8301-11386_3-57433772-76...made-high-tech-materials/ |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
On trouble, yesterday watched a CEP2 job upload. [by coincidence] The _4 file was going at 17Kbs. Gave up after 13 minutes eyeing it, when it was at 50%, but if the desire is to double daily results, you'd be in trouble. This Linux usb 3.0 stick based host is only allowed to have 1 of 8 possible at the time. Everything else WCG was flying, so can't say it was the ISP having me fooled.
--//-- |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Check out this Nature Blog article! Enjoy and share.
http://blogs.nature.com/news/2012/05/us-mater...ative-gains-momentum.html |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7850 Status: Offline Project Badges: 
|
On trouble, yesterday watched a CEP2 job upload. [by coincidence] The _4 file was going at 17Kbs. Gave up after 13 minutes eyeing it, when it was at 50%, but if the desire is to double daily results, you'd be in trouble. This Linux usb 3.0 stick based host is only allowed to have 1 of 8 possible at the time. Everything else WCG was flying, so can't say it was the ISP having me fooled. --//-- On a few occasions I have also watched the big CEP2 uploads and they are consistently slower than the other uploads. I had always thought maybe the ISP was throttling the upload due to its size, but based on what you are seeing perhaps I am wrong.  Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
astroWX
Advanced Cruncher USA Joined: Sep 1, 2007 Post Count: 56 Status: Offline Project Badges: 
|
For what it's worth --
----------------------------------------My primary project is CPDN and it has some comm. files which are substantially larger than anything I've encountered here. I have the high-end of locally available DSL and uploads are sometimes as slow as 7KB/s, other times as fast as 85KB/s. At best, some large files go up in 11+ minutes, other times similar files require nearly a half hour. I've seen downloads as slow as the teens and (rarely) as fast as 750KB/s. (That's for comm. between the US Pacific Northwest and Uni. Oxford in the U.K.) Time of day? Maybe. ISP? Perhaps. However the packets bounce through a lot of hands enroute and every handler has a chance to use its sticky fingers. Servers at Oxford, thanks to a typically underfunded uni. project, are not infrequently hammered into lethargy. So it goes. Of course, sometimes slow up-/down-loads originate in my own house, when my boxes step on each other competing for the resource. Diagnosis, in my estimation, would require far more resources to track-down than I can afford (or care to operate). Better to grin and bear it or, (choose your own metaphor). [Edited for typo.] [Edit 1 times, last edit by astroWX at May 18, 2012 4:54:10 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Note that there may be some interruptions today as the IBM team does some website and forum database maintenance starting at 14:00 UTC.
http://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,33171 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi everybody,
we haven't detected any unusually slow uploads on our end and our servers are not nearly at capacity. Maybe there was a bottleneck on the way here or maybe there was a chance spike in uploads at that very moment... But it is good that you keep us posted on this. Please let us know if this problem persists. Best wishes from Your Harvard CEP team |
||
|
|
|