| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Community
Forum: Chat Room
Thread: Is there a page where you can see estimations (number of years) of these WCGrid WUs goals? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 6
|
|
| Author |
|
|
wcgridmember
Advanced Cruncher Joined: Mar 30, 2005 Post Count: 110 Status: Offline Project Badges: 
|
- estimated number of years that would take each project to finish all current available (on the servers) WUs on one (average) thread?
- same as above but for the current average computing capacity of the World Community Grid community? |
|||
|
||||
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7547 Status: Offline Project Badges: 
|
- estimated number of years that would take each project to finish all current available (on the servers) WUs on one (average) thread? - same as above but for the current average computing capacity of the World Community Grid community? To the best of my knowledge, there is no such page. In order for that kind of page to be valid, the following items would need to be known: the number of work units anticipated to be in the project, the computing capacity of an average thread, and the average of the computing capacity on a per project basis. The latter two could be estimated probably pretty well, but the number of work units anticipated for the project would be the big unknown. Generally speaking, except for ARP, the number of anticipated work units for each project is probably unknown even by the project scientists. At least that is the impression I have been given in the past. If anyone else has a handle on this I would be glad to hear what they think. Cheers
Sgt. Joe
----------------------------------------*Minnesota Crunchers* [Edit 1 times, last edit by Sgt.Joe at Sep 2, 2022 11:14:04 PM] |
||
|
|
wcgridmember
Advanced Cruncher Joined: Mar 30, 2005 Post Count: 110 Status: Offline Project Badges:
|
I am wondering: are the WU generated for the projects:
A: automatically in small batches (no manual intervention from the scientists for each particular batch - and they don´t even know how many WU will be generated in total)? B: by big batches that, if (theoretically) there was an unlimited capacity to finish them quickly, the project would be halted for scientists to generate more WU so the projects could develop more? C: None of the above? If it´s A, then I guess it would be more difficult to make those estimations, but if it´s B, then it´s much easier. Note that the estimations I am talking about are not necessarily for the projects to ultimately finish, but for them to be ¨covered¨ by the computing capacity in a way that would make the ¨bottleneck¨ the capacity for scientists/researchers to make more WU. Say, A project that would finish after 5 big batches of WUs, each generated with the manual intervention of researchers, but we were only in one of those 5 batches. The estimations would be not for the whole project (the 5 batches), but only for one of the batches. |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7547 Status: Offline Project Badges:
|
It is not the same for all projects. Some projects such as ARP use the results from existing computed work units to generate new units. This project has a pretty specific number of work units which need to be processed. See some of the posts by Mike Gibson in the ARP fora for the details. He tracks the progress and gives estimates on the length of the project.
----------------------------------------Other projects such as MCM target different forms of cancer. I don't know how many different forms of cancer they have in mind, but have done several so far. I would speculate they have some goal of about how many work units for each type of cancer they target, but they have not shared the details of the scope of their project. Each project has its own way of generating work units based on the nature of their research. As far as the scope of each project, that is determined by the project designers and the vetting people at WCG. I think the definitive answers to your questions are best answered by the researchers themselves. Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
Cyclops
Senior Cruncher Joined: Jun 13, 2022 Post Count: 295 Status: Offline |
- estimated number of years that would take each project to finish all current available (on the servers) WUs on one (average) thread? - same as above but for the current average computing capacity of the World Community Grid community? Hi wcgridmember, For the first question, Sgt. Joe is correct that we are at the mercy of users turning their computers on and off, we have to validate each workunit by sending it to minimum two people and users also come and go, so "a single thread" is not a great unit of measurement in this case. As for the second question, it would require a lot of estimating and could be thrown off by changes to our backend, new hardware, adopting new projects, etc. I am wondering: are the WU generated for the projects: A: automatically in small batches (no manual intervention from the scientists for each particular batch - and they don´t even know how many WU will be generated in total)? B: by big batches that, if (theoretically) there was an unlimited capacity to finish them quickly, the project would be halted for scientists to generate more WU so the projects could develop more? C: None of the above? If it´s A, then I guess it would be more difficult to make those estimations, but if it´s B, then it´s much easier. Note that the estimations I am talking about are not necessarily for the projects to ultimately finish, but for them to be ¨covered¨ by the computing capacity in a way that would make the ¨bottleneck¨ the capacity for scientists/researchers to make more WU. Say, A project that would finish after 5 big batches of WUs, each generated with the manual intervention of researchers, but we were only in one of those 5 batches. The estimations would be not for the whole project (the 5 batches), but only for one of the batches. It's basically different per project, but the batches being downloaded from the researchers servers on an ongoing basis means that it's completely up to them if they want to manually intervene or not. Changes to the application code are harder to implement, but has been done before for MCM1. The data for the current application version is within the researchers control, so our team talks with them on a regular basis to ensure a constant flow of workunits. This also allows us to meet the current capacity of the grid and not overburden them with more uploaded finished WUs than they can reasonably deal with and analyze. |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7547 Status: Offline Project Badges:
|
Thanks for the clarification.
----------------------------------------Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
|