I have been following Folding@home since the early days. At some point they reinvented themselves. New client and server... I thought they would lose a lot of users, however the new server gave better stats, brining more users into the fold. Then talk begun about a BOINC client for FAH, This made it into beta testing. Fairly popular, but the folks at Stanford put the project on the back burner for the Folding@home Petaflop Initiative(FPI). Projects under this, such as the GPU client in 2006, the just released ps3 client, and the SMP mac client (still in beta).
The Trade off
These clients can only do some kinds of calculations. This is where the brilliance of the people on this project shine through. Throughout the history of this group, decisions were made to do things in a way that would offer the best speed for the hardware people have around the house. Prior to the move to alternative machines, there was a move to alternative cores. Like the more recent moves, this was a way of gaining raw computational power for flexibility. Older tools don't go away! While the number of hours PS3 is the hot new way to get stats for FAH are limited, The benefit of that 5-year-old desktop is still important!
If you find this interesting
I encourage you to listen to Vijay Pande speak. This group has a cherasmatic leader, and a great community at the forums site. Bravo guys, you did it again!
No comments:
Post a Comment