| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Help Fight Childhood Cancer Project Forum
Thread: HFCC WU names & project status |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 36
|
|
| Author |
|
|
martin64
Senior Cruncher Germany Joined: May 11, 2009 Post Count: 445 Status: Offline Project Badges: 
|
The WU names in HFCC include the project phase (s2 currently, which is the last phase as I understood it?). Then there is an 8-digit counter, counting up, where my latest WU is 01715020. Does that number mean anything, or - more important - is there any estimation what the last WU number to crunch will be? Anything beyond s2?
----------------------------------------Regards, Martin  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The WU names in HFCC include the project phase (s2 currently, which is the last phase as I understood it?). Then there is an 8-digit counter, counting up, where my latest WU is 01715020. Does that number mean anything, or - more important - is there any estimation what the last WU number to crunch will be? Anything beyond s2? Regards, Martin Sek's chart shows it as 63% complete; I'm not sure how he derived this # but I'm assuming by the avg hour/task * 9 million experiments. There are 3 targets, and S2 is the last target (3 million compounds each). I have no idea what the #s mean or what the last # is. |
||
|
|
martin64
Senior Cruncher Germany Joined: May 11, 2009 Post Count: 445 Status: Offline Project Badges:
|
Sek's chart shows it as 63% complete; I'm not sure how he derived this # but I'm assuming by the avg hour/task * 9 million experiments. There are 3 targets, and S2 is the last target (3 million compounds each). We have been crunching S2 for quite a while. If the 3 targets have 3 million compounds each, how can we only have achieved 63% up to now? Regards, Martin |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Sek's chart shows it as 63% complete; I'm not sure how he derived this # but I'm assuming by the avg hour/task * 9 million experiments. There are 3 targets, and S2 is the last target (3 million compounds each). We have been crunching S2 for quite a while. If the 3 targets have 3 million compounds each, how can we only have achieved 63% up to now? Regards, Martin I'm not sure - that's a Sek question. I believe maybe the S2 WUs have had much longer run times than the previous two targets... |
||
|
|
martin64
Senior Cruncher Germany Joined: May 11, 2009 Post Count: 445 Status: Offline Project Badges:
|
I believe maybe the S2 WUs have had much longer run times than the previous two targets... Hm, if I got it right, the second target s1 took less time than the first one t1. I am just speculating because I have just put my focus on this project, trying to get my first red badge  So I am wondering what would be the "worst case" for badge hunters (i.e. best case for the researchers) in this project. Here is a wild one:The sequence number might be the number of the compound, so it should count to 3 million for each target. I haven't got the precise figures, but this might have been the case with s1 mid February when we started s2. The second number (4-digit) could be the sequence number within a compound, and the last number, as always, the WU distribution. If that speculation turns out to be true, we would have achieved more than 50% of s2 already, within 2 months. That would mean that the end of HFCC could be in less than 2 months, something like early June. And that would mean: No chance for my Ruby.   Maybe one of the techs or the researchers can comment? Regards, Martin |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
the s2 are highly variable - non-deterministic - to compute. I've got them going on same device in ranges from 5 to 20 hours, project average currently stable at ~7 hours per task.
----------------------------------------s2 is expected to run the longest. The scientists declared 50% complete after s1/t1, and 57% when we had 15,996 CPU years on the counter 4 weeks ago http://www.m.chiba-u.ac.jp/class/bioinfor/wcg/e/hfcc_e/news.html
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
martin64
Senior Cruncher Germany Joined: May 11, 2009 Post Count: 445 Status: Offline Project Badges:
|
s2 is expected to run the longest. The scientists declared 50% complete after s1/t1, and 57% when we had 15,996 CPU years on the counter 4 weeks ago Sek, yes, they are at 57% currently, but I haven't seen them saying s1 is finished. This makes me assume that the results they see are way behind what we are actually crunching. Same as in CMD2. Regards, Martin |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
WCG is pretty fast in offloading the data to the scientists, keeping a copy off-line in storage... few months ago a burst of 3rd copies was send for no reply stragglers just to get some batches finished that were good for 300Gb in live data... why off-loading back to the scientists quickly... you don't want this data to only exist in one place.
----------------------------------------HCMD2 same? It's updated weekly by the scientists? From life crunch [% updated daily] to what their stats show there's about a week lag [Friday night it's done]
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
martin64
Senior Cruncher Germany Joined: May 11, 2009 Post Count: 445 Status: Offline Project Badges:
|
WCG is pretty fast in offloading the data to the scientists, keeping a copy off-line in storage... Ok - but where is the confirmation by te researchers that they received the results for s1 back? What I meant (same in CMD2): what we are actually crunching is way ahead of what the researchers get back as confirmed, complete batches. There are 2 dates that are important at the project end phase: 1. The time when all the regular WUs are distributed - important for the badge hunters 2. the time when all the resuls are returned to the researchers - quite obviously important for the researchers My conclusion from what I observe is that we might run out of WUs (i.e. point #1 above) within the next 2 months. Regards, Martin |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
What I meant (same in CMD2): what we are actually crunching is way ahead of what the researchers get back as confirmed, complete batches. There are 2 dates that are important at the project end phase: No, you got it wrong on HCMD2 particularly. It's what the scientists at several projects confirm back to the members that lags with HCMD2 in top spot of confirming their position versus what we crunch and FAAH in second spot. The uploading/off-loading of completed batches is semi-automatic... frequent... you know IBM mechanics are behind this! Anyway, got something else to do now... HFCC schedule is till end of year... the HPF2 schedule is till next year at least... HFCC at 57% 4 weeks for total project I have no reason to consider as completely off.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
|