Should I fold?

A

angrybusdriver

Gawd
Joined
Jul 16, 2003
Messages
684
I've got an old box that I just use for a TV (MSI TV @nywhere).

It's just a PIII 500, so could I do any good by putting it to work folding?

I just found these forums out of curiosity and it sounds cool so maybe I can help out.

Any help would be cool. Thanks guys.

If so, where do I go to get started?
 
Yes.

Check out the FAQ here:

http://www.hardforum.com/showthread.php?s=&threadid=444196

Download appropriate Folding @ Home v4 Client here:

http://www.stanford.edu/group/pandegroup/folding/download.html

If you're running Windows 2K or XP, run the console version of the Folding Client Software as a service using FireDaemon. You can get a free single license copy of FireDaemon here:

http://www.firedaemon.com

Put the console app somewhere on your hard drive and run the FAH4Console.exe file. When it asks for your team number, make sure to enter 33 (team [H]ard|OCP). After you've gone through the initial configuration and the software begins to start processing a protein, close the client and set it up as a service with FireDaemon.

Welcome!
 
i think the real question is why werent you folding in the first place?
 
Originally posted by angrybusdriver
haha.. It's been 16 hours and I've done 14/400 frames of a work unit :D

Go P3 go.
Click to expand...
There are a lot of little faq's that are not pulled together.
First of all you probably have a Tinker. Look for a file called unitinfo.txt and correlate it with psummary .
Now there are three types of Work Unit (WU) under the Folding umbrella - Gromacs, Tinker, and Genome. Gromacs and Tinkers are proteins with deadlines, Genome is a gene without deadlines, it gets credited to project 799.

Gromacs has the code most attuned to "modern processors" and takes advantage of 3DNow and SSE. Whereas Tinkers and Genome do not. Gromacs "wrings" more out of the processor.
Stanford wants Gromacs back fast. Tinkers seem to be so-so and Genome whenever. Gromacs and Tinkers have deadlines listed in psummary, and Due Times listed in unitinfo.txt. The deadline is when they will reissue the protein, after that it's not good for science. The due time is the last time you can turn it in and get credit (points) for the WU.

"Now why the hell is this nut going on about all this crap" you wonder.

Well ASSuming there is some nebulous relationship between the amount of credit given and the value to science, people try for higher credit, all else being equal.

Gromacs proteins on a machine that runs them with optimization get about 30 to 40 points per Ghz of CPU speed per day. Thus a 3 Ghz machine should get about 100 plus points running 24/7, a 500 Mhz machine about 15 to 20 points. ON GROMACS.

Now I thought Tinkers would get about half of that - so I would expect you to get about 7 to 10 points per day on a Tinker. But if you've got a 70 point Tinker it looks like you are getting less than 4 points per day. If you can get Gromacs and optimizations working you could greatly increase your output.

I'm running a Cely 300@450. It just finished a 70 pointer in about 2 weeks. So about 5 points per day, but it got turned off a couple of times for a few hours. It doesn't have optimizations, heck it doesn't have a case!
 
If i remember correctly, it's an Asus P3B-F. The machine runs cool... (about 23c idle) seriously think I could overclock it? lol

I dunno, that's kind of risky. I already consider it a dying computer even though it still runs pretty well. I just don't trust all of its components.

Edit: nevermind, cpu runs hot while on folding (100% usage over many hours!) around 60c.
 
As it is a SSE enabled CPU, it's probably a good idea to add "-advmethods" too... But using the SSE (Gromacs units) brings the temperature up a bit, so this may be a concern...
 
You must log in or register to reply here.
Back
Top