We're in!

Submitted by praseodymium on Mon, 2005-03-28 22:03.

As of now, we are in the top 100. What will the next goal be? The first place? I don't know. But for now:

Congratulations to all folders, and let's party for a few days!

?????

Submitted by kyleca on Mon, 2005-04-04 22:11.

I just got a new WU for no reason (I didn't finish my old one nor pass the deadline)

Don't worry

Submitted by a_rabbit on Tue, 2005-04-05 14:52.

It could have been a bad WU, or downloaded with errors. As long as it doesn't happen all the time, it's nothing to worry about. If it does happen all the time, then you need to check your system's stability.

system instability

Submitted by salsasonic on Tue, 2005-04-05 15:14.

My theory of what the computer is doing is that it is solving a giant math problem. Now on the scale of things, the more advanced math gets, the more complicated it gets. Every once in a while in Trigonometry, I get an ambiguous case or square root of a negative number or something that does not work. So what I think happens is that the computer tries to fold the protein a certain way, but because of the molecular makeup, it's impossible.
Of course, there could be some very specific and explainable reason that I am completely ignorant of, but that is my take on it.

Check this out....

Submitted by salsasonic on Wed, 2005-03-30 13:56.

Now this is unusual right here...
My screenname is in all three columns right now. I am the highest ranked new user, #14 in production, and also on the user milestone list. Just thought that was cool.

type

Submitted by kyleca on Tue, 2005-03-29 23:26.

the protein i'm working on has 2500 (2,500 not 25,000) frames and it uses the gromacs core. what type do u think it is?

I don't know. possibly genet

Submitted by HopeSeekr of xMule on Wed, 2005-03-30 05:36.

I don't know. possibly genetic tho its probably the initial fold for a *simple* protein, OR you have tweaked the adv settings

Different types of folding

Submitted by HopeSeekr of xMule on Tue, 2005-03-29 19:48.

There are afaik 4 different types of folding, each of which take drammatically different amounts of CPU to finish.

The first is the 'regular' folding. This is the one that requires arguably the least amount of processing. You can recognize this by the client saying 0/250000 steps, the 25000 being how complex the protein is to fold.

Each work unit then goes into a type of genetic algorithm in which it is attemptd to fold the protein as fast as possible using heuristics, generally aiming for under-400 steps. The huge number is the maximum number of folding-algorithms your program will try in order to reach this goal. This takes by far the longest in my experience.

The next step is the verification of this. The protein is attempted to be unfolded in exactly 400 'iterations' that it got from the above step, and if it doesn't work, while you still get credit, the original folding work is reprocessed w/ slightly different algorithms. This takes a medium amount of CPU.

The final type is using the gromacs engine and it has something to do w/ actually modeling the protein (like rasturizing it afaik) and it is done after all these. It is also apparently worth the most points per CPU-hour.

-- No sources, comes straight from my experience, please don't attack the messenger, lol.

specs

Submitted by kyleca on Tue, 2005-03-29 18:48.

1.5 Gigahertz processor, 256 mb ram, windows xp

long

Submitted by kyleca on Tue, 2005-03-29 18:00.

I started a WU with 2500 frames and it says i should be finished in 40 days. is this normal or is my computer slow???

hmm

Submitted by praseodymium on Tue, 2005-03-29 18:14.

I think your computer might be slow, or something else is wrong. Note that it are just estimations, and loose predictions in the beginning.

What are your pc specs?

Go us!!!

Submitted by salsasonic on Tue, 2005-03-29 13:50.

OK, so I haven't helped THAT much, but who cares. Hey, has anyone gotten a WU that starts @ 0/25000? My computer finished it faster than the 400 one that most of the other ones are on right now. Just thought it was really weird. It also got one that started @ 0/2500, but that one took longer. What does the fraction mean? I assumed it was the size of the file, but I don't know.

Frames don't matter

Submitted by praseodymium on Tue, 2005-03-29 15:07.

It's indeed the number of frames you folded. It doesn't really matter how many you did, but it can help you calculating how long it will take for the WU to complete (if you know the number of frames and how long one frame takes).
Mostly you get more points for units with few frames (like 400).

Oh, and the number of WUs doesn't matter either: your points do. There's a team above us with double the number of WUs but just a few points (20k or so) more.

0/25000?

Submitted by Sylf on Tue, 2005-03-29 13:58.

25000 is the number of "frames" that you are folding. On that set of frames, structures are much simpler than others that you can fold them in matter of seconds, giving you some feeling that you're just coasting so fast! Just check out how many points you get from those.... very little ;)

congrats!

Submitted by tak on Tue, 2005-03-29 10:59.

You are all kicking ass, congrats on top 100.
Hope to see you in the top 75 soon :D

tak @ The Shroomery

This also means we're on the

Submitted by Sylf on Tue, 2005-03-29 08:18.

This also means we're on the Team Statistics page of the main site.

Now we have far more visibility than ever before!

Whohooo

Submitted by up north on Tue, 2005-03-29 05:28.

Now we need a 100 party (like the FF 1.0 party)

Anyway cool guys lets keep rolling (and post this fact on SFX!)

-- room for rent --

SFX

Submitted by praseodymium on Tue, 2005-03-29 12:33.

I've posted it on SFX but it isn't frontpaged yet (although it was just a quick entry in my blog).

featured project?

Submitted by up north on Wed, 2005-03-30 05:29.

Is this a featured project, and should me_at_worj and or asa not be nofified? It is a big achievment compaired to the other project that use CPU power (competition = more stiff)