We're gonna crush 'em with our Folding power

Come join the... uh... er... fold.

Moderators: just brew it!, farmpuma

We're gonna crush 'em with our Folding power

Postposted on Mon Apr 15, 2002 10:41 am

Well, judging by the stats, even some of the teams in the top 25 look mighty vulnerable in the face of the furious Folding power of the nascent Team TR. We've already got a score of 1615 with 138 processors active in the last week. Some of those teams ranked around 25-30th place have scores of 10K-12K with fewer active processors. I get the sense some of those teams aren't really going full bore any more.

Of course, there are ascendant teams, like Team Google, that we may never catch. I fear for the [H]orde's top ranking.

But it looks like we have a very good shot of laying the smack down in short order on teams like MacNN, MacOS X, Macphilez, and MacAddict4Life. :) We should break into the top 100 within a week, so we'll be showing up on this page:

http://folding.stanford.edu/cgi-bin/searchteamstats

...for comparsion soon.

I may have to get more power wired into my house, though.

Anybody have some mad stats skillz so we can analyze the contest better?
Damage
TR Staff
Gold subscriber
 
 
Posts: 1671
Joined: Wed Dec 26, 2001 7:00 pm
Location: Lee's Summit, Missouri, USA

Postposted on Mon Apr 15, 2002 12:04 pm

What do we need that isn't being reported alraedy?

Right off the bat, I think that a daily summary of WUs completed and score points gained would be nice. That's the best way for an individual to see if their rate of progress is dropping off, or to see when a new CPU kicks in.

I'm thinking along the lines of a sampling program that downloads a fresh copy of the team stats page every hour, and extracts the numbers from the HTML. That could go into another HTML page, or a database.

The advantage of using a database is that there's an almost unlimited number of views possible that way, even after the fact. So if we captured enough data up front, then discovered later on that some calculated statistic was valuable, we could see that calculation even on historical data.

I developed a system that put data into a database and displayed it via HTML, not too long ago. So I could help. Does anybody have experience with gatting data off of websites? Someone who is handy with regular expressions should be able to whip something up.
You are false data.
Speed
Gerbil Elite
 
Posts: 706
Joined: Thu Dec 27, 2001 7:00 pm
Location: Chicago, IL USA

Hmph

Postposted on Mon Apr 15, 2002 1:16 pm

It took me 4-5 days of running in the background, off and on (I do shut down every night, though), and I've finally completed my first Work Unit. But my client doesn't seem able to send the results and, assumedly, get a new Work Unit to complete. It's been attempting to do so for around half an hour, and nothing. That's something that should be fixed by the people who run that show. I assume it's their server having the problems, but then again I can't say for sure.

Annoying.
Ragnar Dan
Gerbil Elder
Silver subscriber
 
 
Posts: 5354
Joined: Sun Jan 20, 2002 7:00 pm

Postposted on Tue Apr 16, 2002 5:57 am

I dunno if I'm gonna crush *anything* right now.

My bp6 has been doing units on the average of 2 per day. ( 1 per cpu ) but today I decided to check, and after 28 hours of CPU time, one of the units I'm working on has gotten 35% into it. :o Woah, I think I'm getting the work units for Speed, and he's prolly getting mine.

Now I guess I gotta fire the laptop up again and hope it stops overheating.
DiMaestro
Gerbil Elite
 
Posts: 827
Joined: Wed Dec 26, 2001 7:00 pm
Location: North Dakota NoMoah!

Postposted on Tue Apr 16, 2002 6:03 am

Looks like the culprit is proteinAnat. My roomate's computer's been working on one of those for almost a full day ( 1.4 Tbird ). My desktop's got 10 hours of CPU time involved and it's at 47%, ( 1 ghz classic ).

So, if you see your boxen doing one of these, don't despair, it's a toughie.
DiMaestro
Gerbil Elite
 
Posts: 827
Joined: Wed Dec 26, 2001 7:00 pm
Location: North Dakota NoMoah!

Re: Hmph

Postposted on Tue Apr 16, 2002 6:04 am

Ragnar Dan wrote:It took me 4-5 days of running in the background, off and on (I do shut down every night, though), and I've finally completed my first Work Unit. But my client doesn't seem able to send the results and, assumedly, get a new Work Unit to complete. It's been attempting to do so for around half an hour, and nothing. That's something that should be fixed by the people who run that show. I assume it's their server having the problems, but then again I can't say for sure.

Annoying.


Yea, they have had intermittant server problems in the past, and probably into the future as well. When the server goes down, just be patient, it'll come back up, and you'll be able to continue on another work unit.
DiMaestro
Gerbil Elite
 
Posts: 827
Joined: Wed Dec 26, 2001 7:00 pm
Location: North Dakota NoMoah!

Postposted on Tue Apr 16, 2002 8:00 am

DiMaestro wrote:My bp6 has been doing units on the average of 2 per day. ( 1 per cpu ) but today I decided to check, and after 28 hours of CPU time, one of the units I'm working on has gotten 35% into it. :o Woah, I think I'm getting the work units for Speed, and he's prolly getting mine.

You're right on the money, DiMaestro. For the longest time I've been getting WUs from projects 601 and 602. My MPX 1900+ box was putting away a couple dozen of those a day all by itself. You've been getting the 100s, 300s and 500s which take a lot longer to crunch. FWIW, your WUs count for a lot more credits than mine. :)

Ironically, I didn't start getting any of the longer WUs until after I started adding older CPUs. My unscientific research indicates that as far as your score goes, all CPUs are somewhat equal. So if you have an old Pentium laptop, it's well worth adding it to the cause.

I'd sure like to see the distribution of project numbers added into the stats somewhere, so we could get a better idea of how they affect the more obvious numbers.
You are false data.
Speed
Gerbil Elite
 
Posts: 706
Joined: Thu Dec 27, 2001 7:00 pm
Location: Chicago, IL USA


Return to TR Distributed Computing Effort

Who is online

Users browsing this forum: No registered users and 1 guest