| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Influenza Antiviral Drug Search
Thread: Many thanks - Phase 1 now completed! |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 44
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Wow, that will be a fast project. We hope so, since we want to accelerate the drug discovery process and find new drug candidates ever more rapidly ... Stan  |
||
|
|
martin64
Senior Cruncher Germany Joined: May 11, 2009 Post Count: 445 Status: Offline Project Badges: 
|
Wow, that will be a fast project. We hope so, since we want to accelerate the drug discovery process and find new drug candidates ever more rapidly ... Ah, so you could speed up your reearch if the crunching ends earlier? I think your assumption of 30 CPUyears/day is a bit pessimistic, because: 1. At an assumed daily average of 280 CPUyears per day and 8 active projects the average per project should be 35 CPUyears per day 2. Knowing that it will be a short project, badge hunters will jump on it and eat up any work they can get And maybe WCG can tweak the schedulers to put priority on these short-run projects? It will not really hurt the other projects, because they will get a bigger share after the short ones finish. In traffic theory, people call that scheduling method "shortest job first"  Regards, Martin  |
||
|
|
nly32630
Advanced Cruncher The Netherlands Joined: Dec 3, 2007 Post Count: 138 Status: Offline Project Badges: 
|
Wow, that will be a fast project. We hope so, since we want to accelerate the drug discovery process and find new drug candidates ever more rapidly ... Stan The badge hunters will focus on this project, so it might run faster than expected  How will the new badge look like? Who creates this badge and who is deciding to select the right one? I'm looking forward to crunch the phase 2 of IADS (And DDDT2 as well)
Greetings,
Arjan  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
And maybe WCG can tweak the schedulers to put priority on these short-run projects? I do not agree that either short or new projects should get priority. Why should FA@H or HCC get penalized because they take lots of crunching? 30 or 35 CPU years/day; no big difference. What Stan gave us is a relative assessment of the project length which is that is a short project. |
||
|
|
martin64
Senior Cruncher Germany Joined: May 11, 2009 Post Count: 445 Status: Offline Project Badges:
|
I do not agree that either short or new projects should get priority. Why should FA@H or HCC get penalized because they take lots of crunching? You are right with FAAH, as I know that they are scheduling single experiments, and they make use of the results as they are available, without having to wait before the entire project finishes. I don't know that much about HCC etc - if they make use of interim results as well, my statement was nonsense, of course. However, if you have projects that need the entire project's results in order to make any use of it, the picture is somewhat different. A simple example: consider 2 projects, one taking 2 years CPU time, the other 1 year CPU time, and you have one CPU to crunch the projects. What would you do? Have them both run in parallel, giving project one 2/3 and project two 1/3 of the priority? This would be "fair queueing", and both results would be available after 3 years, at the same time. Does it make sense? No. There is another flavour of fair queueing, giving each project an equal share of the resources (which is more or less what you can observe in WCG). This results in an improvement: one project ready after 2 years, the other one taking 3 years still. If you did it with "shortest job first", you would have one result after one year already, and the other one still at the same time as before, i.e. after 3 years. That's the optimum in terms of cumulated calculation time. Regards, Martin [Edit 1 times, last edit by martin64 at Nov 6, 2009 4:50:19 PM] |
||
|
|
skarm
Cruncher Canada Joined: Aug 30, 2009 Post Count: 17 Status: Offline Project Badges:
|
While the above it is quite true it would be unprofessional for anyone at WCG to change priorities even based on mathmatical principle. That would, truthfully, be playing favorites. Also, who is to say that Influenza research should take priority over AIDS research, which in turn should take priority over Cancer research and so forth.
All the projects are very important and individual users will decide themselves which project is the most important. Such as due to my grandfather passing away and living a more troubled life due to Diabetes, I would happily run Diabetes research most of the time if there was such a project active. Regardless DDDT-II and IADS-II will be completed quickly just due to badgehunters and the small amount of WUs. I just do not believe it is WCG's place to dictate which project has more value. |
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 2990 Status: Offline Project Badges: 
|
What you've got to consider (and, I don't know as to whether this'll come into play with DDDT-2 & IADS2), is that both the first phases were 'throttled back'. Obviously, this was for a reason... As to whether that reason will be there for the second 2 phases, time will tell....
---------------------------------------- |
||
|
|
skarm
Cruncher Canada Joined: Aug 30, 2009 Post Count: 17 Status: Offline Project Badges:
|
Well, additionally I suppose the badge requirements could change for DDDT-2 and IADS-2. With such a short project requiring 2 years for Sapphire is most likely unobtainable since many people will be grabbing as many WU as possible.
|
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 2990 Status: Offline Project Badges:
|
Well, additionally I suppose the badge requirements could change for DDDT-2 and IADS-2. With such a short project requiring 2 years for Sapphire is most likely unobtainable since many people will be grabbing as many WU as possible. True - with my 1 dual-core machine, I'll be fortunate to reach Gold in both projects, especially if they start within a matter of days/weeks of each other.... |
||
|
|
skarm
Cruncher Canada Joined: Aug 30, 2009 Post Count: 17 Status: Offline Project Badges:
|
Well, additionally I suppose the badge requirements could change for DDDT-2 and IADS-2. With such a short project requiring 2 years for Sapphire is most likely unobtainable since many people will be grabbing as many WU as possible. True - with my 1 dual-core machine, I'll be fortunate to reach Gold in both projects, especially if they start within a matter of days/weeks of each other.... I have two quad cores going almost full time now barring when my brother is using his for Flight Sim and an older dual core I have to set up, but even with that if I grab 9 days of WUs on each computer I'm still not getting past Gold. |
||
|
|
|