| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Discovering Dengue Drugs - Together - Phase 2 Forum
Thread: DDDT2 Project Recognition Thread |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 857
|
|
| Author |
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges: 
|
I have all of my Windows machines dedicated to DDDT2, but over 60% of the time they are getting WUs from all of the other projects that I have over 20 years contibution on. I want more DDDT2! :) The way DDDT2 WUs are loaded** in the feeder there are frequent short times where there is no WU at all, or some which are waiting for a wingman of another platform. If you have checked the box "If no work is available etc..." and your client asks for work during these short times you will get a WU from another project. If you have not checked the box it will ask again until it gets some, and it will get them sooner or later, and faster than it will crunch them, because there is currently work to distribute for several days, it's not a lie. ** Don't forget that there is this famous 15-MByte file to load first, and download next, therefore these WUs cannot be handled as simply and freely as the common lightweight ones of the other projects. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Next stats update will bring me GOLD!!!
|
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
** Don't forget that there is this famous 15-MByte file to load first, and download next, therefore these WUs cannot be handled as simply and freely as the common lightweight ones of the other projects. This 15.92MB file only is downloaded once per work assignment i.e. if granted multiple DDDT2 tasks it only fetched once (previously also observed by others, similar to HPF2 and RICE). The current mix of CEP2/DDDT2/HFCC is sending about 1 in 10 of DDDT2 to me quad. Because there are multiple active researches selected, there are no red line warnings or back-offs if there are no DDDT2 in the hopper at moment of port-call. It's a slow but definitely wet drizzle. Yesterday 37,705 validated slightly down from the 38,500 estimate. With a bit of push we'll be hitting half a million results validated by tomorrow morning. The standings are now Results Returned 460,373 Crunching on... (whilst creeping up on CEP2 gold too :P)
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
sk..
Master Cruncher http://s17.rimg.info/ccb5d62bd3e856cc0d1df9b0ee2f7f6a.gif Joined: Mar 22, 2007 Post Count: 2324 Status: Offline Project Badges: 
|
Jean, my Q8400 runs out of DDDT2 tasks if I dont give it some attention now and again, but it gets through them faster than my Q6600 and my Phenom II 940, which require less attention. The slower computers tend to build up a healthy cache much easier, I don't need to do anything for my 2GHz dual core. My i7 with HT on the other hand gets through some tasks in 0.36h (21.6min). It's only got a healthy cache because it was running other projects (with longer run times) and had a few days cache. Up until now it has been running mixed tasks from 4 projects. On faster i7's the same 21.6min tasks would only take about 15min, so those systems could get through 32 or 48 tasks in an hour (the communication deferment time), and then pick up 3 tasks before seeing a red message and waiting another hour. As the despatching system has a random element, the system is ok for most people, most of the time, but not all people all of the time.
----------------------------------------So far I have picked up 1909 tasks. With 244days behind me, 690 IP and 274 PV, going Green by the time the flood ends is starting to look possible. If it rains for 2 more days I'd say Green is more likely than not. - I see the completed tasks are starting to clear from the Result Status page. [Edit 1 times, last edit by skgiven at Aug 21, 2010 12:20:02 PM] |
||
|
|
kskjold
Senior Cruncher Norway Joined: May 20, 2008 Post Count: 469 Status: Offline Project Badges: 
|
Silver at last....
---------------------------------------- |
||
|
|
mclaver
Veteran Cruncher Joined: Dec 19, 2005 Post Count: 566 Status: Offline Project Badges: 
|
mclaver, you could create a profile for your DDDT2 machines and choose not to run any other tasks until the DDDT2 project dries up, in a few days or so, then change the profile back to pick up other tasks. Since I do not cache, I do not want to take the chance that my machines will sit idle. I do want to contribute, I am concerned that if I turn off the option to select work from ohter projects if no work available, I will have machines sitting idle. I build machines as a hobby, from parts, and when I am done, I run them 24x7 on WCG. I have 20 windows machines and 5 Linux maches. These range from an AMD 6000 to an I7 975. I normally contribute over a 100 days of time per day, and I try to spread it around to all projects. Right now CEP2 is limited to Unix/Linux machines only and DDDT2 does not have "non-stop" work units. :) I also do significant contribution to other projects (GPUGRID, Milkyway, Collatz, SETI) with my extensive collection of GPUs using CUDA. If WCG supported CUDA, I would change all of this power to WCG and be able to donate 2 years of run time per day to WCG due to the signicanly more power of GPU over CPU. But WCG support of CUDA is another story. :)    |
||
|
|
sk..
Master Cruncher http://s17.rimg.info/ccb5d62bd3e856cc0d1df9b0ee2f7f6a.gif Joined: Mar 22, 2007 Post Count: 2324 Status: Offline Project Badges:
|
I understand your cache predicament, as I also run GPU tasks and have several systems to look after.
----------------------------------------You could still consider creating such a profile for your Windows systems. Then within Boinc select No New GPU tasks, Uncheck the following WCG online profile box: "If there is no work available for my computer for the projects I have selected above, please send me work from another project", Raise your cache temporarily to do a manual update on your windows systems, to force collect DDDT2 work, Then put your normal cache level back, and tick the "box" again. That way you would get more DDDT2 tasks than normal, and hopefully move more hastily towards a green badge. This is pretty much what I am doing for my GPU systems. It's a real Time Kill exercise though, and easy to mess up, hence my request for separation of GPU and CPU configurations within Boinc. Another 4 tasks closer to green: 21/08/2010 18:08:17 World Community Grid Scheduler request completed: got 4 new tasks - Noticed my i7 prefers to run HCMD2 tasks with a deadline of the 31st than DDDT2 with a report deadline of the 29th. So even when you get them you might not be running them. [Edit 2 times, last edit by skgiven at Aug 22, 2010 9:49:53 AM] |
||
|
|
Randzo
Senior Cruncher Slovakia Joined: Jan 10, 2008 Post Count: 339 Status: Offline Project Badges: 
|
Silver mosquito.
Hope with this badge they won`t bite me anymore :-) |
||
|
|
Somervillejudson@netscape.net
Veteran Cruncher USA Joined: May 16, 2008 Post Count: 1065 Status: Offline Project Badges:
|
Ruby probably on Tuesday
|
||
|
|
FAHE
Advanced Cruncher Australia Joined: Apr 27, 2007 Post Count: 122 Status: Offline Project Badges:
|
Silver has appeared. Another 4 days of rain will ensure Gold.
----------------------------------------  |
||
|
|
|