Visitors: 195335
Today: 56
Statistics

204 registered members

Latest member:
wwinter86

More statistics

Mini Poll
This year, you'll be investing in...












Results

Shoutbox

jimerickson: Congratulations rhavern!

rhavern: Quad G34s rock. Mine puts out ~600kPPD.

MarkAGR: Uh? Just woke up from my winter hibernation ... Good Morning everyone! I nearly made it 23m over the winter! I think there's a quad cpu G34 machine on it's way.

toTOW: Anyone alive ?

toTOW: Happy new year to all fellow folders :)

warmon6: Look like the web site needs an update or 2. starting to see cob webs. (could at least mention about bigadv change happening in Jan. ;)

toTOW: Recredit has ben run ... all points should show up now :)

toTOW: Stats are down since the last network outage :(

MarkAGR: OK, so there's a 16 core minimum ... so does anyone know how to produce a 16 core virtual machine from a cluster of ubuntu boxes?

MarkAGR: Where did all thw stats go?

jimerickson: Http://bit.ly/tkpFnJ

jimerickson: 16 core minimum for bigadv. wow!

Adanorm: Hi ! We just applied patches to the site, if shoutbox goes mad, just CTRL+F5 !

jimerickson: Http://bit.ly/okqvf7

jimerickson: Happily folding smp now. and currently earning 2000ppd more than with bigadv. go figure.

jimerickson: I detest p2684, after this one is finished i am moving to smp.

Amaruk: FahCore 11 (ATI) support is scheduled to end September 1st. http://en.fah-addict.net/news/news.php?id=352

hootis: >toTow I think i saw it somewhere either on the folding forum or here, but i cant remember. just wondering if any1 knew.

toTOW: Divery> yes, a little ... on an i7 920 @3.5 GHz (no GPU), I get something like 15k PPD with regular SMP and 22k on BigAdv ...

toTOW: Hootis> did we mention it in one of our news ? anyway I don't remember :(

divery4eyes: Am thinking of adding a couple of smp boxen. is big adv still preferable over regular smp.

hootis: Dose anyone know when the ATi Gpu2 clients are going to be phased out?

MarkAGR: Sniff sniff :(

KaySL: That might explain the weirdly low point yield I'm now getting...

jimerickson: Bigadv bonus reduced from 50% to 20%


Refresh

Archives

Social networks
Facebook
Twitter
Support
­

­

 
Rss Guha Jayachandran
Today, I'll highlight the work of Dr. Guha Jayachandran, PhD. Guha has been a member of the FAH team for several years. He has a undergraduate, masters's, and PhD all in Computer Science from Stanford. He joined the FAH team as a masters student and at that time was heavily involved in revamping the FAH infrastructure. He's been the primary maintainer of the server code and client code.


Guha Jayachandran



For his PhD, his research involved two key areas in the group: studying protein folding and developing new methods for computational drug design. You can see his papers on our site (http://folding.stanford.edu/papers.html) and there are too many for me to describe here, but I'll pick one.

Specifically, I want to talk about "Parallelized Over Parts Computation of Absolute Binding Free energy". Free energies tell us how strong a drug will bind to its target. We want strong binding in order to minimize the dosage given to patients and to improve the specificity (if it binds strong to the target, we can give just a little bit of the drug and so it won't have random side interactions with proteins that aren't the target, which leads to side effects of drugs). Guha's new method allows for a dramatic speed increase in free energy methods, allow us to run calculations in days on FAH which would normally take months in FAH (and potentially decades on regular machines). We are now applying these drug design methods in a variety of applications in FAH.

Copied from: Meet FAH team member, Guha Jayachandran (Vijay Pande's blog)
 
 
Note: No note
(0 note)
Written by: jmn, On: 09/09/09