new projects p3445-3450

The PS3 client has been deprecated.

Moderators: skumar, nrimon, Site Moderators, PandeGroup

new projects p3445-3450

Postby vvoelz » Mon Feb 02, 2009 6:37 pm

I'm pleased to announce the release of several new and exciting projects for the PlayStation3 F@H client. A full description can be found here: http://fah-web.stanford.edu/cgi-bin/fahproject?p=3445

Fs peptide:
project 3445: 251 points, timeout 3 days, deadline 6 days
project 3446: 255 points, timeout 3 days, deadline 6 days

proG peptide:
project 3447: 244 points, timeout 3 days, deadline 6 days
project 3448: 247 points, timeout 3 days, deadline 6 days

Ubq peptide:
project 3449: 255 points, timeout 3 days, deadline 6 days
project 3450: 257 points, timeout 3 days, deadline 6 days

Each WU should complete in about 6.5 hours.

Thanks again for all your work and continued dedication to the Folding@Home project.

Vincent Voelz,
Pande Lab
User avatar
vvoelz
Pande Group Member
 
Posts: 297
Joined: Sun Dec 02, 2007 8:07 pm
Location: Temple University, Philadelphia PA

Re: new projects p3445-3450

Postby J_Rivers » Tue Feb 03, 2009 5:37 pm

It's always a very exciting time for me whenever new projects are unveiled. I've already completed one of the new projects & I'm currently working on a 3445.

Thanks again,

J_Rivers
Image
User avatar
J_Rivers
 
Posts: 101
Joined: Thu Dec 20, 2007 12:06 am

Re: new projects p3445-3450

Postby vvoelz » Tue Feb 03, 2009 5:47 pm

Glad you like the new_projects, J_Rivers. Even in just a few weeks, the data we can collect on these peptides should be really useful to the entire simulation field.
--Vince
User avatar
vvoelz
Pande Group Member
 
Posts: 297
Joined: Sun Dec 02, 2007 8:07 pm
Location: Temple University, Philadelphia PA

Re: new projects p3445-3450

Postby Evanguy » Wed Feb 04, 2009 7:35 am

Great news! Always fun to crunch new projects. Got in and saw a 3445, and was like, "wow, haven't seen that one yet!" and checked the website - pretty neat. They seem pretty important, and it's good to hear these'll be very useful for the whole field. Good job guys.
Evanguy
 
Posts: 25
Joined: Tue Jul 22, 2008 6:34 am
Location: Montreal, Canada

Re: new projects p3445-3450

Postby spazzychalk » Thu Feb 05, 2009 11:24 pm

am i understanding that this project is to test accuracy because all the old gpu and ps3 units are so fast theyre inaccurate? so what about all the other stuff ive turned in in the past?
spazzychalk
 
Posts: 361
Joined: Sun Nov 16, 2008 6:41 am

Re: new projects p3445-3450

Postby codysluder » Fri Feb 06, 2009 2:47 am

spazzychalk wrote:am i understanding that this project is to test accuracy because all the old gpu and ps3 units are so fast theyre inaccurate? so what about all the other stuff ive turned in in the past?


Inaccurate? Where did you come up with that?

Every project is supposed to answer a specific question. As questions get answered, more questions arise. If they've determined how a particular peptide works at a specific temperature and at a specific concentration, the researcher might decide they needed to know what happened at a slightly different temperature or at a different concentration -- and even that is assuming that more study of the exact same protein is warranted.
codysluder
 
Posts: 2128
Joined: Sun Dec 02, 2007 12:43 pm

Re: new projects p3445-3450

Postby vvoelz » Fri Feb 06, 2009 4:22 am

spazzychalk wrote:am i understanding that this project is to test accuracy because all the old gpu and ps3 units are so fast theyre inaccurate? so what about all the other stuff ive turned in in the past?


No, to be clear, all the work you've completed is good stuff!

The issues we're testing are more general. In general, one needs different kinds of models for different purposes -- for example, we use a classical mechanical model for proteins (rather than solving the quantum equations), because it gives accurate results for the time scales in which we're interested. The new projects are testing state-of-the-art models for simulations with implicit solvent. We hope to learn more about the subtle differences between various models that may be revealed now that we can simulate out to such long time scales.

Vince
User avatar
vvoelz
Pande Group Member
 
Posts: 297
Joined: Sun Dec 02, 2007 8:07 pm
Location: Temple University, Philadelphia PA

Re: new projects p3445-3450

Postby spazzychalk » Sat Feb 07, 2009 5:02 am

codysluder wrote:Inaccurate? Where did you come up with that?


The short answer is this: we need to develop even more accurate models of protein folding, so we can best use our models to understand the molecular causes of disease

The long answer is this: Next-generation simulation platforms (like GPUs and the Sony Cell processor) can achieve huge speed-ups in calculation times, at the expense of some physical accuracy.
spazzychalk
 
Posts: 361
Joined: Sun Nov 16, 2008 6:41 am

Re: new projects p3445-3450

Postby codysluder » Sat Feb 07, 2009 9:46 pm

Well, the way I understand scientific calculations, all calculations are approximations and it takes longer to get more accurate results than to get less accurate ones, so part the setup for a new project includes asking the questions "How accurately do you need to know that?" and "How quickly do you need to know that?" and deciding on the proper compromise.

I objected to your word "inaccurate" since there is a wide range of simulations models which are "accurate enough" and none of them fall off the other end of the scale of being inaccurate.
codysluder
 
Posts: 2128
Joined: Sun Dec 02, 2007 12:43 pm

Re: new projects p3445-3450

Postby Amaruk » Sun Feb 08, 2009 9:04 am

I'm with codysluder. Making current/future simulations more accurate does not automatically render past work 'inaccurate'.

For example, Joe Walsh's Maserati does 185. He sends it to a mechanic who's really good and when Joe gets it back it does 200. Does that mean 185 is slow?

Consider pi. It's used to calculate the circumference of a circle. It's also an unending decimal. Here it is taken to 50 places:

3.14159265358979323846264338327950288419716939937510

If we take it to about 12 places (3.141592653589) the circumference of the earth (24,901.55 miles) can be calculated to well within a millimeter. So it could be 40,075,160.083 meters. Or it might be 40,075,160.084 meters. Does taking pi to only 12 places really make that calculation inaccurate?


Just some food for thought. :wink:

Amaruk
Image
User avatar
Amaruk
 
Posts: 481
Joined: Fri Jun 20, 2008 3:57 am
Location: Watching from the Woods

Re: new projects p3451-3458

Postby vvoelz » Sat Mar 21, 2009 3:29 pm

More projects have been released for the PlayStation3 F@H client. Description found here: http://fah-web.stanford.edu/cgi-bin/fahproject?p=3451

Fs peptide:
project 3451: 251 points, timeout 3 days, deadline 6 days
project 3452: 255 points, timeout 3 days, deadline 6 days
project 3453: 255 points, timeout 3 days, deadline 6 days
project 3454: 255 points, timeout 3 days, deadline 6 days

proG peptide:
project 3455: 244 points, timeout 3 days, deadline 6 days
project 3456: 247 points, timeout 3 days, deadline 6 days

Ubq peptide:
project 3457: 255 points, timeout 3 days, deadline 6 days
project 3458: 257 points, timeout 3 days, deadline 6 days

Each WU should complete in about 6.5 hours.

Vincent Voelz,
Pande Lab
User avatar
vvoelz
Pande Group Member
 
Posts: 297
Joined: Sun Dec 02, 2007 8:07 pm
Location: Temple University, Philadelphia PA

Re: new projects p3445-3450

Postby Evanguy » Mon Mar 23, 2009 7:33 am

I have crunched a few of these new WUs - 3450-58, but when I check my WU stats, and all the projects I've done, and how many, I still only see projects 3445-3450. I'm still getting the points and stuff, but it doesn't seem to be tallying them up on the stats page. :?
Evanguy
 
Posts: 25
Joined: Tue Jul 22, 2008 6:34 am
Location: Montreal, Canada

Re: new projects p3445-3450

Postby bruce » Mon Mar 23, 2009 8:30 pm

Evanguy wrote:I have crunched a few of these new WUs - 3450-58, but when I check my WU stats, and all the projects I've done, and how many, I still only see projects 3445-3450. I'm still getting the points and stuff, but it doesn't seem to be tallying them up on the stats page. :?
See the News item:
http://folding.typepad.com/news/2009/03 ... ended.html
bruce
 
Posts: 21670
Joined: Thu Nov 29, 2007 10:13 pm
Location: So. Cal.

Re: new projects p3445-3450

Postby SavantStrike » Mon Mar 23, 2009 10:29 pm

I hadn't seen that news item... interesting..
Folding 24x7 IHS
Team 31630. Nerds For Jesus. Join us, we could use the help! Or check us out at nerdsforjesus.org
SavantStrike
 
Posts: 71
Joined: Wed Oct 29, 2008 2:14 pm

Re: new projects p3445-3450

Postby Evanguy » Tue Mar 24, 2009 7:57 am

bruce wrote:
Evanguy wrote:I have crunched a few of these new WUs - 3450-58, but when I check my WU stats, and all the projects I've done, and how many, I still only see projects 3445-3450. I'm still getting the points and stuff, but it doesn't seem to be tallying them up on the stats page. :?
See the News item:
http://folding.typepad.com/news/2009/03 ... ended.html

I should really start checking the news page more often! :oops:

Thanks for that! Seems like a good deal, I remember when it was nearly impossible to get any stats a while back, this is a good change until there's a permanent fix for slow updates.
Evanguy
 
Posts: 25
Joined: Tue Jul 22, 2008 6:34 am
Location: Montreal, Canada

Next

Return to PlayStation3 client

Who is online

Users browsing this forum: No registered users and 1 guest

cron