|
Written by Team 45
|
|
Friday, 02 March 2007 14:37 |
Q: What is this Folding@home anyway?
A: Folding@home is a distributed computing project that uses our
computers to unravel the mysteries of protein folding. We are helping
scientists at Stanford University to understand why and how proteins
'fold', so it will be easier for them to find cures for diseases. When
proteins do not fold correctly (i.e. "misfold"), there can be serious
consequences, including many well known diseases, such as Alzheimer's,
Mad Cow (BSE), CJD, ALS, Huntington's, Parkinson's disease, the
influenza virus, Osteogenesis Imperfecta, and many Cancers and
cancer-related syndromes. The successes and discoveries of the project
can be found in part on this page: Stanford Protein Folding Research
Q: How big is the project?
On February 18, 2009, Folding@Home passed the 5 petaflop mark, making
this project the most powerful distributed computing cluster in the
world by far, bar none. To put it in context, traditional
supercomputers have just broken the 1 petaflop mark, and even that
level of performance is very challenging to aggregate. Millions of
machines have donated cycles to the project since its inception.
Q: This sounds interesting, can I help?
A: YES! You sure can, and it is free too! All you need to do is
download a small application, and let it run in the background on your
computer. It will only make use of the unused clock cycles, so you will
never notice that it is running. Once in a while it will connect to the
internet to get new info (a new work unit) to process. The folding of a
work unit (WU) takes several hours to days, depending on what project
and what client you are running, and the processing speed of your
computer.
Q: Minimum system requirements?
A: Minimum system requirements are currently a ~ 200MHz pentium MMX
computer folding 24/7. The faster the computer, the more work you can
return. If your computer does not live up to the minimum requirements,
it will not be able to complete the work units within the deadline. If
this is the case, we suggest you look into other DC projects instead
that don't have a work returning deadline.
Q:
I know there are f@h clients available for Linux, Mac OS X, and Windows
operating systems. Can Sony Playstation 3 consoles fold?
A: Yes! The PS3 makes for a very powerful folding@home computer. Find out more by reading Stanford's PS3 FAQ.
Q: Can GPUs run a Folding@home client?
A: Sure, both ATI GPUs and NVidia GPUs are supported. Further information on this in the ATI FAQ.
Q: Where are the FAQs?
A: Here: Folding Client FAQs
Q: Which version should I choose?
A: If you are on a Windows machine, you can choose from several
clients. The standard client is stable and will run on a wide variety
of hardware. The SMP client is optimized to use multi-core machines:
dual, quad, and more. There are graphical and console versions of most
clients. The console versions are faster for folding, but do not show
fancy graphics. There are standard and SMP versions available for
Windows, Mac, and Linux/BSD (for Linux/BSD SMP folding, a 64-bit setup
is required).
Q: Where can I get the client?
A: Here: Client Downloads
Q: Won't Folding@home slow up my computer?
A: Folding@home is set to only engage the unused clock cycles in your CPU, so if the CPU
is doing something else, Folding@home steps back. This means that your
performance drops by running Folding@home are absolutely minimal to
none. If the computer isn't being used for anything else, Folding@home
will take over. However, GPU folding has been found to slow up graphics
on some computers, so it isn't recommended to play games, for example,
while folding. It's better to shut down the client while doing GPU
intensive operations.
Q: Folding@home and overclocking?
A: Most of us on team 45 are overclockers with highly overclocked
computers. The more MHz, the faster a work unit is processed. However,
it is vital to the scientific aspect that your overclocked computer is
completely stable. Folding@home will stress your CPU
with activity and heat more than any software stress tester will,
including Prime95 and OCCT, and an unstable computer may crash. We
suggest you run various stability testers (like the Prime95/OCCT
torture tests) before using an overclocked computer for Folding@home.
Q: Will Folding@home fry my CPU?
A: No, but Folding@home stresses the CPU
quite a lot. It is important that you have sufficiently cooled your
computer. This is something standard users should not have to worry
about. With that being said, computers today are MADE for work. There
is no proof suggesting that Folding@home has ever caused a CPU to fry, nor is there any evidence claiming that it will shorten the life of your CPU.
Q: Does my computer have to be online all the time to fold?
A: No, your computer only needs a connection to the internet when
returning and retrieving a work unit. How often depends very much on
your computer, but a couple of times a day for about 2 minutes isn't
unlikely. You can fold even if you are using a modem - just set it up
to auto dial and auto disconnect and Folding@home will take care of the
rest.
Q: Can I add more than one computer to the same user account?
A: Yes, no problem. You can install the Folding@home program on as many
computers as you like. Just make sure to use the exact same username
(case sensitive) and team number on all of them.
Q: I would like some additional info, where should I go?
A: You are always free to ask any questions in this forum topic. The official Folding@home site is also worth a look: Stanford Folding Homepage
Also, there is a F@H forum here: Folding Community Forum
Vijay Pande has a blog here that he updates regularly: Vijay's Blog
Q: Great, I've downloaded and installed the application, now what?
A: Now you are ready to join our team and fold! Run the application,
and you will be questioned with username, team number etc. Username is
up to you, and as team number just enter number 45 and you will
automatically join our team, AOA Forums.
Q: Is there a guide to installing the client software?
A: Yes, there are guides for every client here: Official Stanford Install Guides
You may also ask for help in our forums. There is a forum for SMP as
well as one for GPU folding clients. You will find lots of user
experiences there that will help you to get going.
Q: So what is all this I hear about how much fun it is? Is there a competition?
A: While the real aim of the project is and always will be the science
and the service to humanity, it's clear that a little friendly
competition can go a long way to keeping donor interest in the project.
The Pande team has designed a method of benchmarking the work in their
own lab, and releasing them with an attached point value that you earn
each time you complete a job. The points are regularly tallied for both
individual donors and teams, and are published here: Official Stanford Folding Stats
There are also independently run stats pages, and our favorite is here at extremeoverclocking.com: Team 45 User Standings
Q: What is a work unit?
A: A work unit is a small work packet that the client will download and
work on.It consists of an amount of "frames" (typically 100 frames) and
when all frames have been folded, the client will return the results to
the server and fetch a new work unit.
Q: What is the difference between work units and points/credit?
A: There are many different kinds of work units, some take much longer
time to process than others. Because of this each unit is assigned to a
certain value of points, so you will recieve more credit for a "big"
work unit than you will for a smaller work unit.
Q: What is this about deadlines for returning work units?
A: All units are based on the results of previously returned units, so
it is important to get the results back to Stanford as soon as
possible. For this very reason, you need to return the work packets
within the given deadline for them to be of any use. If you fail to do
so, your work is a waste. WUs for the standard client have a 1 or 2
month deadline, SMP WUs currently have 6 days, and GPU WUs are 2-4 days.
Q: What determines the speed of my folding client?
A: A rule of thumb is that Folding@home is all about raw CPU speed. RAM speed does not seem to have much impact on folding times, so a fast CPU is just about all that counts for SMP folding. For GPU folding, it's the graphics card that rules
.
Q: I want to fold FAST. Which CPU/GPU should I buy?
A: Dual or multi-core CPUs a great for folding. They allow you to have
more than one instance of the f@h client running. In the case of a
dual-core, you could potentially double your output, compared to a
single core CPU.
Q: Are there any advanced flags I can use when running Folding?
A: Yes, there are a few. You may want to add "-advmethods" (without
quotation marks) to your Folding@home short cut to gain access to
non-released units. These units have been through beta testing and just
need the final touch up before being released to the public. We suggest
you do indeed add this flag, as you are likely to encounter newer units
and thus receive more credit. "-verbosity 9" will make the client more
chatty and helps a lot in troubleshooting. We recommend that the client
be run with it at all times. There are others in the SMP FAQ.
Q: Anything else I should know?
A: Yes, once you have joined the active ranks of the folding team, feel
free to add our official folding badge to your signature. Read this
thread, and follow these instructions to do so: Team 45 Badgel
You'll need to know what your Folding@Home username is, so make a note of it if you don't already know.
Next, click here to edit your signature
Once you're ready, add the following line, but remember to change the
"PUT_YOUR_ID_HERE" so that it is EXACTLY your Folding@Home username!
PHP Code:
PHP Code:
[team=fah]PUT_YOUR_ID_HERE[/team]
It will look like this:
Q: No, really, is there anything else?
A: Look at and read the sticky threads in the AOA Folding forum. They
will direct you to other useful links, and interesting stuff about the
project.
WELCOME to our team, thank you for joining - I hope you enjoy folding with us.
Regards,
Allan Nielsen
Team Captain (former)
This FAQ has been shamelessly edited and updated by ThunderRd (9-09).
|
|
|
