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: 10
[ Jump to Last Post ]
Post new Thread
Author
Previous Thread This topic has been viewed 1313 times and has 9 replies Next Thread
Johnny Cool
Ace Cruncher
USA
Joined: Jul 28, 2005
Post Count: 8621
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Longer work units and variables

I don't think it's just me. Seems like when I do run my PC 24/7 (I don't do it that much, as my PC is a biz mosheen), the work units appear to run much longer.

Just curious!

Thanks!
----------------------------------------

Team Andrax Co-Captain
Free-DC Stats
Join Team Andrax at WCG
[Feb 4, 2006 3:43:23 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Longer work units and variables

I saw great variation in the duration of Human Proteome Folding tasks. One of my machines got three tasks which took 48 hours each (1.4 GHz machine running 24/7). Others were done in 5 hours.

Fight Aids At Home tasks all seem to complete in about 10 hours. 1 hour per 10%. But again, that's on my 1.4 GHz machines. Surprisingly consistent.
[Feb 4, 2006 3:56:57 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Longer work units and variables

I'm relatively new, only been running part time weekdays and full week ends. The first few units took 40 hours on my spair P 233 machine - Now the units are taking 60- 120 hrs. On my P11 400 I'm at 75% and 65 hours run time First few units were 18 - 24. On the P4 2.58 Units would pop off at 3-6 hours - Now looking at 40- 75 per unit.

Not alone in the increase. I thought I had read they were trying to have a unit take about 40 clock hours per each type machine. So they may readjust - will see in the future
[Feb 4, 2006 2:19:57 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Longer work units and variables

C320 --

The objective is for the work units to take about 10 hours for the average user. Most of the time I have found they do a good job at this. I have noticed that the FightAIDS@Home seem more consistant as noted above by rutgers-nieer-cgilbert. Over the 14 months of the Human Proteome Folding project, the work units have seemed to vary quite a bit. There may be some aspect of the Rosetta processing that makes it difficult to keep within the guidelines. At the moment, I have noted that several people are saying they have "monster" or long running work units.

Best Regards,
[Feb 4, 2006 3:34:32 PM]   Link   Report threatening or abusive post: please login first  Go to top 
coolstream
Senior Cruncher
SCOTLAND
Joined: Nov 8, 2005
Post Count: 475
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Longer work units and variables

Several P4s running here and one old P2 just for the fun of it.

In the past, I noted fondly that the old P2 would soldier on for several days just to give one result whilst the bigger powerful machines could sometimes achieve 2-3 results per day (only Folding selected).

In the past few days, I am seeing these machines too working for 50-90 hours per result. I check them often and sometimes see one has manged to grab a 'short' unit which it crunches successfully in a couple of hours and then it's back to another monster. devilish

With the amount of machines I am running, each 6-hour period usually yields a few days of processing time on the 'My Devices' page. This has now changed to being a sporadic pattern.

All things considered, I am basically saying that a machine that would crunch 2-3 units per day is with these newer units managing one and a bit per day, and slower machines are now taking me up to a week to process a unit.

It's not a problem for me because I have seen this happen before both here and at UD and I am still well within the allocated periods for units. For people new to here and to the project, I would say to hang tight and not worry. Smaller units will come back wink
----------------------------------------

Crunching in memory of my Mum PEGGY, cousin ROPPA and Aunt AUDREY.
[Feb 4, 2006 7:46:37 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Johnny Cool
Ace Cruncher
USA
Joined: Jul 28, 2005
Post Count: 8621
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Longer work units and variables

Several P4s running here and one old P2 just for the fun of it.

In the past, I noted fondly that the old P2 would soldier on for several days just to give one result whilst the bigger powerful machines could sometimes achieve 2-3 results per day (only Folding selected).

In the past few days, I am seeing these machines too working for 50-90 hours per result. I check them often and sometimes see one has manged to grab a 'short' unit which it crunches successfully in a couple of hours and then it's back to another monster. devilish

With the amount of machines I am running, each 6-hour period usually yields a few days of processing time on the 'My Devices' page. This has now changed to being a sporadic pattern.

All things considered, I am basically saying that a machine that would crunch 2-3 units per day is with these newer units managing one and a bit per day, and slower machines are now taking me up to a week to process a unit.

It's not a problem for me because I have seen this happen before both here and at UD and I am still well within the allocated periods for units. For people new to here and to the project, I would say to hang tight and not worry. Smaller units will come back wink


Smaller units will come back? I most certainly hope so. It would be great to send in a work unit every day.

Let's hope that WCG realizes this as well. Running *one* business PC 24/7 is difficult for me.

I am giving it all that I can and would like to see shorter work units.
----------------------------------------

Team Andrax Co-Captain
Free-DC Stats
Join Team Andrax at WCG
----------------------------------------
[Edit 2 times, last edit by Johnny Cool at Feb 7, 2006 4:06:00 AM]
[Feb 7, 2006 3:05:08 AM]   Link   Report threatening or abusive post: please login first  Go to top 
keithhenry
Ace Cruncher
Senile old farts of the world ....uh.....uh..... nevermind
Joined: Nov 18, 2004
Post Count: 18665
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Longer work units and variables

After over a year of running WCG using both the UD agent and BOINC, I've found a surprising amount of consistency with the FAAH WUs, relatively speaking. It appears that far more often than not, they do in fact run within 8-12 hours. HPF WUs tend to have had a good bit more variability. Most complete with one if not two days. I've see a number run in a matter of hours and have gone as long as almost five days on a single WU. The main thing in every case is to let 'em crunch as they will finish, one way or another.
----------------------------------------
Join/Website/IMODB



[Feb 7, 2006 12:43:45 PM]   Link   Report threatening or abusive post: please login first  Go to top 
owentnz
Cruncher
Joined: Sep 18, 2005
Post Count: 8
Status: Offline
Reply to this Post  Reply with Quote 
Re: Longer work units and variables

I've been running long WU's. The longest was 5 days & 10 hrs. That's on a Pentium III 450MHz with a processor score of 42.
[Apr 14, 2006 8:08:43 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Robert Tusler
Cruncher
Joined: Apr 4, 2005
Post Count: 27
Status: Offline
Reply to this Post  Reply with Quote 
Re: Longer work units and variables

This pentium 4 with a processor score of 187 was known to take over 80 hours for a work unit under hpf. My record of work units done shows fewer work units than average for the amount of points earned. I assume that in some way the server knows which are the big mippers and saves the big work units for them.
[Apr 16, 2006 5:59:27 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Longer work units and variables

Hello Robert Tusler,
The server hands out work units randomly. Something else is involved or this is just chance.

Lawrence
[Apr 16, 2006 7:52:32 AM]   Link   Report threatening or abusive post: please login first  Go to top 
[ Jump to Last Post ]
Post new Thread