 |
 |
possible o have one ype of F@H unit
|
|
|
|
|
Senior User
Join Date: Mar 2003
Status:
Offline
|
|
is it possible to only get gromacs instead of tinker as tinker os like 133% SLOWER than gromacs on my G5?
|
MacbookPro dual 2Ghz 1GB Ram 128 Graphics
|
| |
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Apr 2002
Status:
Offline
|
|
If you are using the terminal when you start f@h use ./osx-3.25 -advmethods Using this I have only had one tinker unit in the last month.
|
|
|
| |
|
|
|
|
|
|
|
Senior User
Join Date: Oct 2002
Location: God's Country, The South
Status:
Offline
|
|
Originally posted by Kevin Moon:
If you are using the terminal when you start f@h use ./osx-3.25 -advmethods Using this I have only had one tinker unit in the last month.
I have had more than one Tinker unit lately on mine. It took so long to process that when I see I have one now, I throw it out and get a new one. My old Dual G4/533 is just Tooooo slow on those units! It took two whole weeks to process one 70 point Tinker unit. It can do a 51 point Gromacs unit in about 3 days, way too much difference.
If you stop your client and throw out the work folder, then restart it, it will get a new work unit automatically. If it is a Tinker unit too, repeat the process until happy. 
|
|
|
| |
|
|
|
|
|
|
|
Forum Regular
Join Date: Mar 2002
Location: 42 minutes from the other side of the world
Status:
Offline
|
|
Does that mess up the data for the project?
|
|
|
| |
|
|
|
|
|
|
|
Senior User
Join Date: Oct 2002
Location: God's Country, The South
Status:
Offline
|
|
Originally posted by jarling:
Does that mess up the data for the project?
I assume that if a WU is not turned in within some specified time frame, it is re-issued. Most DC projects issue more than one copy of every WU just in cast that happens.
|
|
|
| |
|
|
|
|
|
|
|
Forum Regular
Join Date: Mar 2002
Location: 42 minutes from the other side of the world
Status:
Offline
|
|
But this project cant re-issue work because what it sends out is time sensitive. That is, it is based on previous results some number of nanoseconds earlier in the folding process.
Once the servers send out a certain number of units for a given protein, on a given folding pathway, at a given time along that pathway, thats it. Right?
Or do they go back and re-issue from the middle of such a chain if they find the results to be unreliable, or not replicated enough?
|
|
|
| |
|
|
|
|
|
|
|
Senior User
Join Date: Oct 2002
Location: God's Country, The South
Status:
Offline
|
|
Originally posted by jarling:
But this project cant re-issue work because what it sends out is time sensitive. That is, it is based on previous results some number of nanoseconds earlier in the folding process.
Once the servers send out a certain number of units for a given protein, on a given folding pathway, at a given time along that pathway, thats it. Right?
Or do they go back and re-issue from the middle of such a chain if they find the results to be unreliable, or not replicated enough?
Based on the fact that some number of units will never be returned for various reasons, such as computer crashes, there must be some allowance made for re-issuing WUs that are not returned in a reasonable amount of time. If not, at some point it would all tumble down.
|
|
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
 
|
|
|
Forum Rules
|
|
|
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
|
HTML code is Off
|
|
|
|
|
|
|
|
|
|
|
|
|
Top
