Index  | Recent Threads  | Unanswered Threads  | Who's Active  | Guidelines  | Search
 

Quick Go »
No member browsing this thread
Thread Status: Active
Total posts in this thread: 53
Posts: 53   Pages: 6   [ Previous Page | 1 2 3 4 5 6 | Next Page ]
[ Jump to Last Post ]
Post new Thread
Author
Previous Thread This topic has been viewed 16347 times and has 52 replies Next Thread
Steve WCG
Senior Cruncher
Joined: May 4, 2009
Post Count: 216
Status: Offline
Reply to this Post  Reply with Quote 
Re: The Clean Energy Project - Crunching Chart

Show the server you are a reliable quick returner.

Not relevant to beta ... they want to test all machines, not just good ones. If your PC wants a WU and there is BETA up on the queue, as long as you have BETA selected, that is what you will get.
[Aug 16, 2009 11:12:25 AM]   Link   Report threatening or abusive post: please login first  Go to top 
roundup
Veteran Cruncher
Switzerland
Joined: Jul 25, 2006
Post Count: 843
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: The Clean Energy Project - Crunching Chart

Updated extrapolation:

73,52% aug 05
74,73% aug 08
75,05% aug 09
75,94% aug 12
76,36% aug 14.

This makes 2,84% in 9 days, that is 0,316% per day
This pace would give us another 75 days for CEP.
The average is now a little lower, so the estimated time went up a few days smile

Remember this is just extrapolation, figures can change

This extrapolation is not valid anymore as uplinger explained in this thread:
http://www.worldcommunitygrid.org/forums/wcg/viewthread?thread=26577 :
As I mentioned I would get a better estimate for everyone. Currently we are sitting at 24 days of work waiting to be loaded into BOINC. There is about 2 days of work loaded into BOINC on a normal day. This means the estimate is around 26 days of new work for CEP1.

Also, a note on Sek's charts...His estimates were pretty close if we were completing the work units that caused a majority of the error 29's a few weeks ago. This caused his estimate to be off target as members have noticed.

Hope this helps,
-Uplinger

This was written on Aug 20. According to the quoted post we should have about 22 days left today (Aug 24) including the WU's that have to be loaded into BOINC.
To verify this I try a new extrapolation:
In Sek's Crunching Chart we saw the following figures:
Aug 20 86,42
Aug 23 87,54
This is 1,12% in 3 days or 0,3733% per day.
At this pace we had about 33 days left which is far away from uplingers outlook.
Or - as an reverse calculation - the figures in the crunching charts will be at about 95,75 % in the 22 days that are predicted by uplinger. Maybe the remaining 4,75% percents will be repair units and a inaccuracy factor?
I think we should calculate with 97,75% instead of 100% in order to get more reliable extrapolations, shouldn't we?
[Aug 24, 2009 4:32:37 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Steve WCG
Senior Cruncher
Joined: May 4, 2009
Post Count: 216
Status: Offline
Reply to this Post  Reply with Quote 
Re: The Clean Energy Project - Crunching Chart

Seeing as there are a fixed number of WUs left and there are a fixed number of WUs distributed each day, there is no extrapolation necessary. It is only a matter of counting the days as provided by uplinger.
[Aug 24, 2009 4:37:50 PM]   Link   Report threatening or abusive post: please login first  Go to top 
roundup
Veteran Cruncher
Switzerland
Joined: Jul 25, 2006
Post Count: 843
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: The Clean Energy Project - Crunching Chart

Seeing as there are a fixed number of WUs left and there are a fixed number of WUs distributed each day, there is no extrapolation necessary. It is only a matter of counting the days as provided by uplinger.

True. Nevertheless it may be interesting to see how the Crunching Chart fits to uplinger's prediction.
[Aug 24, 2009 5:00:17 PM]   Link   Report threatening or abusive post: please login first  Go to top 
uplinger
Former World Community Grid Tech
Joined: May 23, 2005
Post Count: 3952
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: The Clean Energy Project - Crunching Chart

roundup,

Please note that all of my estimates are for new work loaded into the grid. This means that all of the work units that are waiting to be sent to member computers have been sent.

Sek will have to tell you what his charts and percentages represent. But in my mind they represent the actual completion of work units and this is about 1-2 week behind when a work unit was sent out.

So, from my latest estimates we are about 21.5 days away from running out of new work that is sent to the members computers. Please note that the project usually runs for about 2 weeks finishing and cleaning up tasks.

Hope this helps,
-Uplinger

Edit: Latest estimate is actually at 21.1 days...
----------------------------------------
[Edit 1 times, last edit by uplinger at Aug 24, 2009 5:38:30 PM]
[Aug 24, 2009 5:32:18 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Sekerob
Ace Cruncher
Joined: Jul 24, 2005
Post Count: 20043
Status: Offline
Reply to this Post  Reply with Quote 
Re: The Clean Energy Project - Crunching Chart

The % progress is not tuned to the amount of work handed out. That is way way ahead of what comes back and for that uplinger's days are when the 'new' work issue is depleted. Secondly, make sure that you compare % of the same timestamp. The 87.54% was the (est.) noon value i.e. could it have been 1.12% on 2.5 days? Certainly no chart got refreshed to show the full end-of-Sunday position given that the cumulatives are including a substantial spill due a delay of stats-start.

Anyway, I'll leave you to it guessing the day the exact last result will come back for phase 1. For DDDT that took approximately 18-20 days, yes really. On July 25 the validated tasks total started dipping and on Aug, 12+13 each there was a single result validated either from a straggler or a rerun of a result in the techlab.

edit: See that uplinger posted meantime while typing... he's pretty close to what I'm doing.

The DDDT validated results history as a guide:

Jul.20-09 53419
Jul.21-09 54359
Jul.22-09 53336
Jul.23-09 54115
Jul.24-09 51676
Jul.25-09 29546
Jul.26-09 17664
Jul.27-09 15545
Jul.28-09 12396
Jul.29-09 9081
Jul.30-09 7246
Jul.31-09 5997
Aug.01-09 4497
Aug.02-09 3483
Aug.03-09 3192
Aug.04-09 1239
Aug.05-09 439
Aug.06-09 238
Aug.07-09 128
Aug.08-09 55
Aug.09-09 24
Aug.10-09 7
Aug.11-09 10
Aug.12-09 1
Aug.13-09 1
----------------------------------------
WCG Global & Research > Make Proposal Help: Start Here!
Please help to make the Forums an enjoyable experience for All!
[Aug 24, 2009 5:44:27 PM]   Link   Report threatening or abusive post: please login first  Go to top 
nly32630
Advanced Cruncher
The Netherlands
Joined: Dec 3, 2007
Post Count: 138
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: The Clean Energy Project - Crunching Chart


To verify this I try a new extrapolation:
In Sek's Crunching Chart we saw the following figures:
Aug 20 86,42
Aug 23 87,54
This is 1,12% in 3 days or 0,3733% per day.
At this pace we had about 33 days left which is far away from uplingers outlook.
Or - as an reverse calculation - the figures in the crunching charts will be at about 95,75 % in the 22 days that are predicted by uplinger. Maybe the remaining 4,75% percents will be repair units and a inaccuracy factor?
I think we should calculate with 97,75% instead of 100% in order to get more reliable extrapolations, shouldn't we?


Thank you for your calculations.

Like Steve WCG said, the end date must me fixed with the controlled maximum number of WU's per day.
I think the main reason why your extrapolation is going to 95,75% is that there are many WU's being processed OR on pending validation. Those WU's are not validated yet.

When the last WU is sent out, it will probably take 14 days before everything is finished. 10 days for the WU's and the WU's that did not get returned then have another 4 days as a repair unit :-)
Assuming all last repair WU's will be returned in 4 days...

Edit: I wrote this message before I read the comments of Uplinger and Sekerob. They already wrote better answers than I did. :-)
----------------------------------------
Greetings,
Arjan

----------------------------------------
[Edit 1 times, last edit by nly32630 at Aug 24, 2009 7:55:15 PM]
[Aug 24, 2009 7:51:00 PM]   Link   Report threatening or abusive post: please login first  Go to top 
roundup
Veteran Cruncher
Switzerland
Joined: Jul 25, 2006
Post Count: 843
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: The Clean Energy Project - Crunching Chart

roundup,

Please note that all of my estimates are for new work loaded into the grid. This means that all of the work units that are waiting to be sent to member computers have been sent.

Sek will have to tell you what his charts and percentages represent. But in my mind they represent the actual completion of work units and this is about 1-2 week behind when a work unit was sent out.

So, from my latest estimates we are about 21.5 days away from running out of new work that is sent to the members computers. Please note that the project usually runs for about 2 weeks finishing and cleaning up tasks.

Hope this helps,
-Uplinger

Edit: Latest estimate is actually at 21.1 days...


uplinger and Sekerob,
thanks for your explainations. This is very useful to understand that your different figures are caused in different prerequisites for your calculations. This helps for future extrapolations in other projects!
[Aug 25, 2009 6:57:09 AM]   Link   Report threatening or abusive post: please login first  Go to top 
nly32630
Advanced Cruncher
The Netherlands
Joined: Dec 3, 2007
Post Count: 138
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: The Clean Energy Project - Crunching Chart


uplinger and Sekerob,
thanks for your explainations. This is very useful to understand that your different figures are caused in different prerequisites for your calculations. This helps for future extrapolations in other projects!


Roundup, this is still interesting, I agree.

Uplinger said there is a delay of 1-2 weeks between sending out and completed results.
What about calculating 2 weeks?
So: Extrapolate the progress and subtract 2 weeks to guess the end of sending out WU's.

If I calculated well, with the details from Uplinger, the last normal WU's would be sent out on sept 15 :-)
----------------------------------------
Greetings,
Arjan

[Aug 25, 2009 8:20:41 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Sekerob
Ace Cruncher
Joined: Jul 24, 2005
Post Count: 20043
Status: Offline
Reply to this Post  Reply with Quote 
Re: The Clean Energy Project - Crunching Chart

You probably did not see the exchange in a parallel CEP thread allowing for even greater refinement ;>)
----------------------------------------
WCG Global & Research > Make Proposal Help: Start Here!
Please help to make the Forums an enjoyable experience for All!
[Aug 25, 2009 8:32:21 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Posts: 53   Pages: 6   [ Previous Page | 1 2 3 4 5 6 | Next Page ]
[ Jump to Last Post ]
Post new Thread