/sci/ - Science & Math

>>2041517

Get /v/ in on it; I guarantee none of them are actually using their PS3s.

good idea

What's this folding about?

>>2041526
proteins.

>>2041531
Link? I'd like to get involved.

http://folding.stanford.edu/

If you have a dual core computer, find the appropriate "SMP" client under the "high performance clients" page. Same for the GPU client if you have a capable graphics card.

Doesn't have to be a dual, btw...a tri/quad(or more) can do it, and a pentium 4 with Hyperthreading can do it as well. Otherwise, run the unicore client on the main page.

Here's some guides to get you set up. Not that the 4chan team is # 47633, so put that instead of OCN's number.

SMP client tutorial:
http://www.overclock.net/folding-home-guides-tutorials/682249-windows-7-setting-up-smp-client.html

GPU tutorial:
http://www.overclock.net/folding-home-guides-tutorials/683706-windows-7-setting-up-gpu2-gpu3.html

Multi GPU tutorial:
http://www.overclock.net/folding-home-guides-tutorials/578934-windows-7-multigpu-guide.html

>>2041517
I like having the full performance of my rig at my disposal at all times.

Wait, so is there some kind of program to open? All I'm getting is a process. I didn't read the description thoroughly... what exactly is the purpose?

>>2041559
You hand over your some of your processing power to be used by labs round the world that run simulations on protein folding.

>>2041550
It's on low priority...you don't lose any performance. When you start up a game, F@H's process is like "yessamassa, I go an' make you some chickin'".

>>2041559
Yerp. You just let it run. Kind of like being part of a botnet, except you know it's happening, and it's actually doing something useful.

bump

>>2041573
Oh, so I'm helping them with their research? That's fucking awesome.

>>2041559

It's a distributed computing project. This one in particular is studying, at its core, how proteins get their shape, and what factors affect this. Research from it will go to help numerous causes, mainly research on diseases like Cancer, AIDS, and the like.

I think some people do it just for the points, though...to be competitive. Like a game. And 4chan has the manpower to be #1, so I don't see why they shouldn't, you know?

>>2041574
low priority uses some of the processor still. I'm greedy, cpu/gpu usage>other people's scientific endeavors for me.

So it only works if my computer is on, right? Because I leave my computer on 12/7 instead of 24/7.

>>2041586
Are you actually asking this question?

There's a 4chon team?

Need a new HDD before I begin folding again

>>2041586

...yes, haha. But that's ok. Just set tit to start up when windows does, and it'll be fine. It doesn't require constant usage. It doesn't even need constant internet access when it's running...just when it's turning in results and getting a new work unit.

>>2041584

Because you're competing against the likes of Hardforum and EVGA

LOL

>>2041599
EVGA FTW!

>>2041585

You misunderstand. Low priority means it still has 100% cpu usage, but when something else wants to use the cpu, this steps aside and lets it.

It doesn't really interfere with anything, even when it's running. the only negative effect it could possible have on your computer is if your cooling sucks and it gets hot.

eh how often do they update the user list? i dont see mine yet. also, do i have to be connected to the internet?

>>2041605

Well the site I linked isn't their official one, it's just the best one for stats...so I have no idea. The actual folding site at stanford.edu updates it every few hours, I think.

>>2041604
You let that run at 100% usage? Damn, folding@home is gay.

>>2041605

As soon as you complete a work unit and upload your results you will appear on their page. It gets updated a couple of times a day.

Worry about processing first! And have the team number correct.

>>2041612
I definitely have the number correct. The usernames refer to memes.

>>2041611 oh noes 100%

This is awesome.

What other applications could this method be used for?

Also, how many people need to run the program to equal a contemporary supercomputer?

>>2041611

You don't HAVE to run it at 100%, but most people do, yes.

>>2041601

FUCK YEAH

>>2041620
I realize that, and know to cap usage.Thing being even though its very hard for shit to max out my rig, who gives a fuck?

When you go to Config > Advanced > CPU usage percent, which way is it? 0 to 100 or the other way? I know this is a dumb question, but there's no indicator...

>>2041623

Pissed off because your computer is too gay?

>>2041619
Well, the original one was SETI@home, SETI being Search for ExraTerrestrial Intelligence. There's a bunch of other projects as well, doing work ranging from climate models to mapping virus outbreaks to studying the stars.

As far as the second question, it depends on the supercomputer.

>>2041637
I stole it from some random person at Starbucks, haven't had time to modify the parts.

>>2041643

nigger

>>2041633

Moving it to the right means more CPU usage, I believe.

You're on the unicore client, right? (The SMP one is console only) If you run the unicore client on a multicore system, it can only max the equivalent of one core. This gets spread out, so on a dual core, the maximum usage for the unicore client is 50%. It drops to 25% on a quad, and so on.

>>2041637
If by gay you mean run laps around your 32bit piece of 90s shit, than yeah. Its six-cores of 64bit gay.

>>2041633
GODDAMMIT JUST UNINSTALL F@H I DONT WANT YOU TO EVEN CONTRIBUTE TO THIS SCIENTIFIC CAUSE JUST FUCKING KILL YOURSELF.

>>2041650

AMD fag detected.

What if I have a dual core laptop, but it's frankly not a high performance dual core laptop?

>>2041648
>You're on the unicore client, right? (The SMP one is console only) If you run the unicore client on a multicore system, it can only max the equivalent of one core. This gets spread out, so on a dual core, the maximum usage for the unicore client is 50%. It drops to 25% on a quad, and so on.

I can't even understand what you're saying. No acronyms please. Also, I just stolet this laptop, so I'm clueless on the core details.

>>2041660
>mfw

>>2041656

The stupidity!

>>2041666
u butthurt, faggot?

>>2041660

Read the FAQ on the folding website. Follow the steps EXACTLY.

>>2041658
Can you guys stop questioning each other's sexuality and answer my question?

>>2041658
That answers your question now doesn't it?

>>2041660
SMP means symmetric multi processing. Basically, anything with two or more cores.

Unicore means one core.

Running the 1-core client can only max out one core. The computer speads the work out, however. So what you end up with is the work equivalent to one core being maxed being spread out among all of the cores, but since there's only enough work for one core, then they can't all be used to their full potential. Thus, on a dual core, you can't go higher than 50% usage because there isn't enough work for more than that.

>>2041674

It will work.

ANY laptop is more prone to overheating so keeping an eye on the temps is important before letting it loose at 24/7

>>2041673
>All science and math related topics welcome.
Homework threads will be deleted, and the poster banned.
>No "religion vs. science" threads.
>Follow the steps EXACTLY.
Wait what?

>>2041682

Folding website you fucking troll

>>2041679
So that means dual cores are faster, right?

>>2041681
But if I stay on the single core client, and it's actually just running at 25% on each core, then I don't NEED to worry about overheating, right? So doesn't it actually make more sense for me to stay on the single core, as that way I can contribute (less, admittedly) without endangering my hardware even a little bit?

>>2041686
Oh! I didn't read your post carefully. How silly of me.

>>2041686
So..what's the link to their website? My googlefu is weak.

>>2041687
Yep.

>>2041691
More or less, if you have a quad. 50% if you have a dual.

>>2041692
>>2041693
I don't want to live anymore.

>>2041693
folding.stanford.edu

Yes, that's stanford university.

This page for the high performance clients:
http://folding.stanford.edu/English/DownloadWinOther

>>2041549
^that post for tutorials

hard

>>2041696
Dual. AMD Athlon X2 Dual-Core QL-62 2.00 GHz

>>2041705
Yup, 50% on that.

It'll be 25% if you set the program's CPU usage to 50%, though.

You guys can use HWMonitor to monitor your temperatures:
http://www.cpuid.com/softwares/hwmonitor.html

As long as nothing tops ~70C or so, you're good. Graphics cards can get hotter, but I have a feeling most people won't run the GPU client.

This thread needs to be archived.

>>2041708
On an ATI Raden HD 3200, what would be the point?

I know fully well this computer isn't some sort of hardcore number.crunching gaming rig.

BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET BOTNET

>>2041708
Yeah, running the single core right now, the cores are at 66°C. I think I'll stick to this.

>>2041708
Is it okay if I put my laptop in the fridge?

>>2041719
Only if you have air filtration in the house that removes all water traces from the atmosphere so that the air is completely dehydrated. Because otherwise, condensation --> fucked computer.

>>2041719
That won't prevent you from being part of the botnet.

>>2041722
Disregard him, he is an idiot, the more water the better, then you get water cooling.

So, yes, yes, put your laptop in the fridge.
/g/ here.

>>2041728
Water isn't the same as mineral oil. OMG please dunk your desktop in a bucket and tell me what happens.

>>2041733
full of water

>>2041728
>/g/ here
Credible enough for me! Gonna get some ice cold water.

>>2041733
WHOOSH.
Fuck me, /sci/ really are idiots.

>>2041733
Nothing happened, what now faggot.

Currently running the SMP2 client on my Core 2 Duo E4500. On the current project I'm getting about 1000ppd. Highest I've seen with this processor was ~1300ppd

>>2041742
turn it on.

I'm folding for 4chan, and been folding for a while now and am part of the 14 active users.

>>2041711

That card isn't supported.

nvidia really kick ass when it comes to CUDA. Anything from the 8000 series above will work with the gpu client.

>>2041747
I used to, then I stopped, but now this thread's reminded me and am currently folding.

20/2000

This could take a while...

Remember that if you're going to do the GPU client make sure you choose a Nvidia card with at least 96 CUDA cores. Radeon support is awful compared to GeForce.

GT 240 related, 96 CUDA cores

It does take a while...I've got an athlon II x4 @ 3.4GHz and it usually takes most of the day to finish a work unit.

>>2041760
Yep, it is...they're working on a new ATI client though.

The problem is the architecture. ATI cards are better at lots of smaller calculations, whereas nvidia is better ar larger, less numerous ones.

>>2041761
I'm jealous that you can complete a WU in less than a day. Takes me about a day and a half or more.

how do you make it run 1 client and run 90-100% on all cores? or do you have to run multiple clients which don't get the same jobs given to them from what I've seen.

takes me 5 days to finish a wu

>>2041765
You run the SMP client, which can use all of your cores. It's console-only, but still relatively easy to set up, and you can use TrayIt to put it into the system tray where it won't bother you.

I have a second monitor, so I just stick it there instead...
<<like so

I have a linux server/router with an Athlon X2 in my closet that mostly downloads torrents so it's quite free to do some other work. Will it work on linux? I'll have to check this out.

>>2041777
>but still relatively easy to set up
Just make sure you follow directions in the F@H page exactly. I glossed over the set up instructions and it would only use 50% cpu power. Make sure you do it right!

http://folding.stanford.edu/English/WinSMPGuide

>>2041783
Yes, it actually works better on linux than on windows.

I think it works on OS/X and BSD as well.

>>2041783
the SMP client will actually run better on Linux (64-bit only however)

GPU will run a little worse.

>>2041765

You can run multiple instances of the unicore and set the affinity. Processing two WU at a time. 100% each core.

>>2041777
What cooler are you using with your Athlon?

i got lost in this long thread somewhere.

i got a quadcore + ht and a great gpu. what client makes sense to get the most out of this machine for tests?

i also got a ps3, can i send packages from multiple systems (i.e. pc + ps3) for the same user acount (aka myself) or do i have to create multiple accounts (one for each system / client software) ?

>sudo emerge -av foldingathome

I love linux.

>>2041804
If you want to max out your system set the SMP client to use 3 of the cores leaving one core open because the GPU client will use the spare core for some work.

Also you dont need multiple accounts for one user name. if you set it up on multiple machines it'll ask you for your passkey, you need one issued to you from the F@H website, however the PS3 doesn't seem to ask for one so you wont need it there.

me = dumb

so i should run two clients, the SMP and a GPU one? there is no app that does all in one?

can you please give me links to the clients? because scrolling through the .edu site with downloads & expert user downloads i just get lost between all possible versions ...

(oh, and: windows 7 64bit here)

>>2041817

SMP2 Client
http://www.stanford.edu/group/pandegroup/folding/release/FAH6.30-win32-SMP.zip

And the directions to set it up are here
http://folding.stanford.edu/English/WinSMPGuide


What GPU do you have?

Do we do anything else in this team except overusing our CPUs?

You know, like weed smoking sessions or whatever?

thanks a lot for the link!

only nvidia gtx 460

OT: captcha: "gov ends PHILOSOPHY", not kidding =)

>>2041796
Cooler Master Hyper 212 plus. I think they're <$30 right now...can't be beat for the price.

It's only at 3.3Ghz right now, but I think it's got more in it.

>>2041847

You whore. Every shop on the citadel is your favourite.

I found the 4chan page on the official website:
http://fah-web.stanford.edu/cgi-bin/main.py?qtype=teampage&teamnum=47633

Looks much better...shows 97 active clients within the last 50 days.

According to this page, 4chan is set to overtake a few teams shortly:
http://folding.extremeoverclocking.com/team_overtake.php?s=&t=47633

It's running but I mostly see it using only one of my cpu's.

And yes, I changed the configuration to make use of two.

http://www.allprojectstats.com/showuser.php?projekt=0&id=2048322

herpa derpa

>>2041899
Changed the configuration how? The SMP client should do that by default.

>>2041903
Nice, what sort of rig are you running?

For anyone wondering what he posted, BOINC is a network of other distributed computing projects. It's like a central program that you can run them all from. F@H isn't part of it, though. I don't see why not...

>>2041905

I edited the CPU variable and changed it's value from 0 to 2 in the /etc/conf.d/foldingathome file.

Whoa, there's a 4chan BOINC team as well:
http://www.allprojectstats.com/showuser.php?projekt=0&id=2048322

the GPU3 fah client that is linked from all places (googling, forums, f@h official size) is on some users private space and this isn't avaialble.

even the howto mentioned on one of the top posts in this thread links to it, and is only 5 days old ...

anybody know why / where else to get gpu3 (beta) ?

>>2041914
You're on linux?

>>2041922

I sure am. Maybe it's one of the advanced options, I'll look around.

I do SETI and Einstein, and sometimes Milky Way. Sorry /sci/.

>>2041919
Somewhere in here, I think:
http://foldingforum.org/viewtopic.php?f=59&t=16471#p163637

>>2041928
Good idea...I honestly only know how to set up the unicore client in linux. :(

yes exactly, the last link works - there you can find GPU3 systray / comandline versions.

thanks =)

I've been running this on my living room pc for about a month. I don't use it, so it seemed good use as any. I don't contribute to society in any other form.

contributing right now with 8547 ppd (7 cores) and 9838 ppd (gpu).

how much is this in comparison to others?

Just switched my team to 4chans, it was 0 which I think is anonymous, hope it helps.

>>2041986
Well, I do 4k total at best, usually in the low 3k's, so I'd say pretty damned good. You'd have to have some huge, power-hungry GPU's to do much more, I think.

I'd like to know what was going on at the beginning of the year...?

Wait...we haven't failed this hard since the end of '08. WTF guys?

What hardware is everyone running their F@H clients on?

Do you know how many PPD you are getting? If you don't you can measure it using the HFM.net tool

http://code.google.com/p/hfm-net/

>>2042655
Mine's an Athlon II x4 635@3.3GHz, and according to HFM I usually get between 3k and 4k depending on what WU I'm on.

I'd like to overclock it a bit more but it's a C2 chip...I haven't even been able to get 3.5 stable, so far. I have badass RAM, though, so that helps a bit. (DDR3 1600@7-8-7@1.6v)

Folding bump