Particularly those who have quad core.

3/19/2009 3:16:42 PM

Ok,

So I'm having a problem. Whenever I leave my (quad-core) computer on at over night....it stops and restarts. I'm running the SMP client. Is this fixable or a common problem?

3/21/2009 8:37:29 AM

Well, I've been folding for well over five years and have yet to lose a single piece of hardware in that time. As long as you maintain cool temps you should be fine (spend $30 for a Xigmatek S1283 CPU cooler & crank up your GPU fan speed).

3/21/2009 10:55:42 AM

Can I change the percentage of resources folding@home uses on my iMac. It's really lagging.

3/26/2009 2:36:53 PM

bttt!

4/29/2009 7:40:52 AM

bump for 2.5MILL



[Edited on May 4, 2009 at 5:55 PM. Reason : pic]

5/4/2009 5:54:33 PM

3 million, time to quit



[Edited on May 13, 2009 at 1:11 AM. Reason : ;]

5/13/2009 1:10:58 AM

don't quit!

6/2/2009 8:51:08 PM

bump per request

9/2/2010 1:18:57 PM

i was thinking about this the other day.

9/2/2010 1:27:51 PM

my PS3 doesn't play games or blue rays anymore, but I guess I could hook it back up so that it'll do some folding.

I'll get my work machine to start folding as well.

9/2/2010 1:31:58 PM

points: 360,324
24hr Avg: 168
Today: 118
WUs: 1261
UserRank: 27062
Team Rank: 36

So when I check it again in a week, I can see how I'm improving. [/dork]

9/3/2010 11:10:59 AM

Uh, you know that's what this webpage is for:

http://folding.extremeoverclocking.com/user_summary.php?s=&u=281961

And who the heck is dowroa? That user just contributed 1,000,000 pts LAST WEEK! At my best month ever was only 300,000 pts. and took me almost 2 years to pull 3,000,000!

http://folding.extremeoverclocking.com/user_summary.php?s=&u=500936


[Edited on September 3, 2010 at 11:50 AM. Reason : .]

9/3/2010 11:45:24 AM

So I haven't updated my computer in a while but i'm ordering new parts soon. Getting an i7 950 with an Nvidia 460, probably getting a second 460 soon after that. Might do some folding. Maybe be able to pull some points with that?

9/3/2010 11:46:50 AM

dowroa is a top 40 folder in the world right now.

9/3/2010 11:51:54 AM

Here's an updated version of the top post:



Support Cancer Research and the NCSU Wolfpack With Your Computer's Spare Time!

What Is It?
Essentially Folding@Home (http://folding.stanford.edu/) uses your unused CPU cycles to model how proteins fold in the human body. The misfolding of proteins is often involved in well known diseases, such as Alzheimer’s, Mad Cow (BSE), CJD, ALS, and Parkinson’s disease. Because this endeavor requires massive amounts of computing power, their project uses spare CPU time on computers that would otherwise sit idle in homes and businesses across the world. Here at North Carolina State University, we have formed this team in order to donate our spare CPU time to help this important endeavor.

The North Carolina State University (http://ncsufolding.syprus.com) Folding@Home Team (Team # 59) was one of the first teams to join the Folding@Home project, and was the very first University-based team to join. We are currently ranked in the top 100 in the world and are ramping up production in an effort to make it into the Top 50. We are all dedicated to doing our part to help end some of these terrible diseases, and we would appreciate the company of anyone who would like to help.

All it takes is the installation of a small piece of software and your computer will start completing Work Units for the project and the team. The Folding@Home process runs at the lowest possible priority so you can run it 24X7 without slowing down your computer or interfering with your everyday computing tasks. If you have a Sony PS3 Game Console you can also run Folding@Home (http://folding.stanford.edu/English/FAQ-PS3) on it with great results.

How Can I Help?

Go to http://folding.stanford.edu/English/Download,download and install the software. Our Team Number is 59.

Click here - http://ncsufolding.syprus.com/?page_id=4 for more detailed installation instructions.

Windows Computers:
We recommend using the "text-only console” client for most windows computers. If you have a dual/quad/etc core processor, be sure to install and run the SMP client to get you and the team far more points than the standard console client (but sometimes it has problems starting as a service, so you might have to restart the client manually or automatically via startup folder when your computer restarts). In general, its best to stay away from the graphical clients. If you have a Pentium 4 with Hyperthreading, its best to turn off Hyperthreading in the BIOS if you know how. Feel free to post in this thread with any questions.

Mac and Linux Computers:
There are both "text-only" consoles and SMP clients for Linux and Macintosh computers too! If you have a dual/quad/etc core processor be sure to run the SMP client as you'll get far more points than the text-only client. But the text-only clients also help out too if you only have single-core processors.

PS3:
PS3s currently provide over 60% of the processing power in the Folding@home network and are capable of producing many points for the team. If you have PS3 system version 1.6 or later, you will see a Folding@Home icon in the Network column of the XMB (PS3 Networking menu). Just click on the icon and that's it. If you don't have 1.6 or later, perform a system upgrade. Make sure to use Team #59!

The more computers and PS3s you run the clients on, the more points your username and the NCSU team earns! Install as many clients as you can and watch your username/team progress at the stats page - http://folding.extremeoverclocking.com/team_summary.php?s=&t=59

Links

Team Links:
NCSU F@H Team Website: http://ncsufolding.syprus.com
NCSU F@H Team Discussion: http://www.thewolfweb.com/message_topic.aspx?topic=118820
NCSU F@H Team Facebook Group: http://ncst.facebook.com/group.php?gid=5012521644

F@H Links:
F@H Project Homepage: http://folding.stanford.edu/
F@H Official Forums: http://forum.folding-community.org
F@H Wikipedia Article: http://en.wikipedia.org/wiki/Folding@home
F@H Wiki Site: http://en.wikipedia.org/wiki/Folding@home

Stats:
Stats1: http://folding.extremeoverclocking.com/team_summary.php?s=&t=59
Stats2: http://folding.stanford.edu/English/Stats
Monitor your production: http://fahmon.net
Compare your production with other users: http://fahinfo.org

[Edited on September 3, 2010 at 12:14 PM. Reason : l]

9/3/2010 12:11:57 PM

neodata686 that pc should be able to pull 30,000ppd

Folding with a 4ghz overclock the processor alone can do 25,000 if all 8 cores are dedicated to the work unit. The gtx460 can do 10,000ppd.

9/3/2010 12:44:25 PM

Crazy production. Passed my like I was standing still. Actually I am. I basically only have laptops now or old outdated desktops that are not worth folding on. I stopped using my PS3 because the fan was running loud even when I was not folding/playing games.

9/3/2010 3:05:23 PM

so here is one of the boxes I have FAH running on:

[root@server ~]# cat /proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      : Intel(R) Xeon(TM) CPU 2.80GHz
stepping        : 9
cpu MHz         : 2793.346
cache size      : 1024 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 1
bogomips        : 5586.69

processor       : 1
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      : Intel(R) Xeon(TM) CPU 2.80GHz
stepping        : 9
cpu MHz         : 2793.346
cache size      : 1024 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 1
bogomips        : 5585.25

processor       : 2
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      : Intel(R) Xeon(TM) CPU 2.80GHz
stepping        : 9
cpu MHz         : 2793.346
cache size      : 1024 KB
physical id     : 3
siblings        : 2
core id         : 0
cpu cores       : 1
bogomips        : 5585.27

processor       : 3
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      : Intel(R) Xeon(TM) CPU 2.80GHz
stepping        : 9
cpu MHz         : 2793.346
cache size      : 1024 KB
physical id     : 3
siblings        : 2
core id         : 0
cpu cores       : 1
bogomips        : 5585.28

9/3/2010 4:44:17 PM

that's debatable, i've heard 1x SMP client per 2 cores, so if you have 4 cores, you could have 2 SMP clients. if you don't run SMP, you could at minimum run 4 non-SMP clients.

9/3/2010 5:55:11 PM

so if I wanted to run 4 non-smp clients on this box, I would need to install it 4 times or just run the program 4 times?

I have it configured to run as a service, so do service folding start x4?

9/4/2010 9:13:23 PM

they just have to be in different folders and they have to have different machine ID's.

c:\fah1\fah.exe -local (machine ID 1)
c:\fah2\fah.exe -local (machine ID 2)
c:\fah3\fah.exe -local (machine ID 3)
c:\fah4\fah.exe -local (machine ID 4)

run fah.exe -configonly and go through the advanced settings to change machine ID.

check task manager to see if all cores are being utilized, if not you may have to manually set each process to a core using Task Manager, or you can download a program called Affinity Changer to do it.

[Edited on September 5, 2010 at 1:31 AM. Reason : .]

9/5/2010 1:24:09 AM

my bad, I forgot to mention earlier, this box in running linux.

9/5/2010 7:34:19 AM

tl:dr - Monster box here is pushing crappy PPD numbers. Linux Live CD, Octo-box, SMP client. I am doing something wrong.


This box is doing just hundreds of PPD...can't be right. I am running the SMP client in a single instance and 'top' shows just under 800% CPU utilization for the FAH process. Should I be using the regular client? Also set to do 'big' workloads.

[root@livecd ~]# cat /proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      :                Intel(R) Xeon(TM) MP CPU 3.66GHz
stepping        : 9
cpu MHz         : 3670.001
cache size      : 1024 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 1
apicid          : 0
fpu             : yes
fpu_exception   : yes
cpuid level     : 5
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush
 dts acpi mmx fxsr sse sse2 ss ht tm syscall lm constant_tsc pni monitor ds_cpl est tm2 cid cx16 xtpr lahf_lm
bogomips        : 7340.00
clflush size    : 64
cache_alignment : 128
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 1
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      :                Intel(R) Xeon(TM) MP CPU 3.66GHz
stepping        : 9
cpu MHz         : 3670.001
cache size      : 1024 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 1
apicid          : 1
fpu             : yes
fpu_exception   : yes
cpuid level     : 5
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush
 dts acpi mmx fxsr sse sse2 ss ht tm syscall lm constant_tsc pni monitor ds_cpl est tm2 cid cx16 xtpr lahf_lm
bogomips        : 7338.40
clflush size    : 64
cache_alignment : 128
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 2
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      :                Intel(R) Xeon(TM) MP CPU 3.66GHz
stepping        : 9
cpu MHz         : 3670.001
cache size      : 1024 KB
physical id     : 3
siblings        : 2
core id         : 0
cpu cores       : 1
apicid          : 6
fpu             : yes
fpu_exception   : yes
cpuid level     : 5
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush
 dts acpi mmx fxsr sse sse2 ss ht tm syscall lm constant_tsc pni monitor ds_cpl est tm2 cid cx16 xtpr lahf_lm
bogomips        : 7338.55
clflush size    : 64
cache_alignment : 128
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 3
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      :                Intel(R) Xeon(TM) MP CPU 3.66GHz
stepping        : 9
cpu MHz         : 3670.001
cache size      : 1024 KB
physical id     : 3
siblings        : 2
core id         : 0
cpu cores       : 1
apicid          : 7
fpu             : yes
fpu_exception   : yes
cpuid level     : 5
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush
 dts acpi mmx fxsr sse sse2 ss ht tm syscall lm constant_tsc pni monitor ds_cpl est tm2 cid cx16 xtpr lahf_lm
bogomips        : 7338.54
clflush size    : 64
cache_alignment : 128
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 4
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      :                Intel(R) Xeon(TM) MP CPU 3.66GHz
stepping        : 9
cpu MHz         : 3670.001
cache size      : 1024 KB
physical id     : 8
siblings        : 2
core id         : 0
cpu cores       : 1
apicid          : 16
fpu             : yes
fpu_exception   : yes
cpuid level     : 5
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush
 dts acpi mmx fxsr sse sse2 ss ht tm syscall lm constant_tsc pni monitor ds_cpl est tm2 cid cx16 xtpr lahf_lm
bogomips        : 7338.50
clflush size    : 64
cache_alignment : 128
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 5
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      :                Intel(R) Xeon(TM) MP CPU 3.66GHz
stepping        : 9
cpu MHz         : 3670.001
cache size      : 1024 KB
physical id     : 8
siblings        : 2
core id         : 0
cpu cores       : 1
apicid          : 17
fpu             : yes
fpu_exception   : yes
cpuid level     : 5
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush
 dts acpi mmx fxsr sse sse2 ss ht tm syscall lm constant_tsc pni monitor ds_cpl est tm2 cid cx16 xtpr lahf_lm
bogomips        : 7338.55
clflush size    : 64
cache_alignment : 128
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 6
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      :                Intel(R) Xeon(TM) MP CPU 3.66GHz
stepping        : 9
cpu MHz         : 3670.001
cache size      : 1024 KB
physical id     : 11
siblings        : 2
core id         : 0
cpu cores       : 1
apicid          : 22
fpu             : yes
fpu_exception   : yes
cpuid level     : 5
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush
 dts acpi mmx fxsr sse sse2 ss ht tm syscall lm constant_tsc pni monitor ds_cpl est tm2 cid cx16 xtpr lahf_lm
bogomips        : 7338.54
clflush size    : 64
cache_alignment : 128
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 7
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      :                Intel(R) Xeon(TM) MP CPU 3.66GHz
stepping        : 9
cpu MHz         : 3670.001
cache size      : 1024 KB
physical id     : 11
siblings        : 2
core id         : 0
cpu cores       : 1
apicid          : 23
fpu             : yes
fpu_exception   : yes
cpuid level     : 5
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush
 dts acpi mmx fxsr sse sse2 ss ht tm syscall lm constant_tsc pni monitor ds_cpl est tm2 cid cx16 xtpr lahf_lm
bogomips        : 7338.54
clflush size    : 64
cache_alignment : 128
address sizes   : 40 bits physical, 48 bits virtual
power management:

9/7/2010 10:32:30 AM

no one has any suggestions for this?

9/10/2010 11:17:58 AM

The big work loads get big points if you turn them in early. Like a i7 folding 24/7 with a 4ghz overclock will get roughly 55,000 points. The down side is it will take 2.5 days to submit it netting (~22,000ppd). If you taking longer to chug out work units then your points will suffer greatly because of the bonus system.

The only linux stuff I tried when folding was using the premade VM Ubuntu that came from the evga.com forums http://www.evga.com/forums/tm.aspx?m=1305 I got it up and running smoothly but i could never dedicate enough time to finish work units reliably since it was a gaming/work computer.

I remember having a dell server with two 3.4Ghz xeons trying to fold a regular smp. I got it started and over the weekend it only got 30 some percent done. It was a dog. I hope thats not your case.

9/12/2010 12:21:15 AM

I'm gonna have to change up my strategery (thanks GW for that word, I love it!).

I've only racked up about 7600 points (around 70 something Work Units) in the last 18 days, and that's with 4 servers.

I had expected a lot more production.

9/20/2010 10:12:11 AM

dowroa put up like 7 million points last month. Holy crap!!

9/20/2010 10:16:03 AM

Well I know who has a slightly higher than normal power bill last month. F-me. Hopefully he posts up what he is using to be the #19 PPD F@H producer in the WORLD.

He could get 300K PPD with 3 systems but they each would have to be SR-2 mobos with dual hexacores. Thats only a minimum of $3000 each... chump change

9/20/2010 8:33:24 PM

yeah I'd like to know his setup as well.

9/20/2010 8:45:45 PM

He has 45 active clients in the last week:
http://fah-web.stanford.edu/cgi-bin/main.py?qtype=userpage&teamnum=59&username=dowroa
I'm not sure if that is for each processor or each install.

9/20/2010 9:35:13 PM

http://www.myspace.com/dowroa ?

9/20/2010 9:44:55 PM

The 45 clients come from each install when you put in machine # into the configuration. If he has a dual socket mobo that would count as one smp client otherwise each cpu and gpu is a client.

9/21/2010 7:58:24 AM

hmm yeah I checked that page with my username, I have some FAIL going on. I currently have 4 servers (see the cat /proc/cpuinfo above), my desktop at home (should probably make that smp), and my laptop here at work.

I guess this weekend I'll be redoing my folding stuff.

9/21/2010 9:15:51 AM

So back when I had my Conroe I believe I ran one SMP client and one console client.

With my new i7 950 and nvidia 460 what should i run? I started up a regular console client last night and it was hardly even utilizing one thread of one core. I need to run one for each core? And each one will utilize two thread? Then one SMP for the 460?

9/21/2010 9:16:58 AM

Go to the Stanford High Performance Page http://folding.stanford.edu/English/DownloadWinOther

Nvidia fermi gpus need gpu3 so download the top link and install it like normal.

Processors need the smp console client, which is the last link. Just double click the .exe file and it will ask you the normal folding questions. There is an install guide if you havent installed the smp client (though I know you have neodata686). Two to note are these:

1) Set -advmethods flag always, requesting new advanced scientific cores and/or work units if available (no/yes) [yes]

2) Additional client parameters [] -smp 7 or -smp 8 (7cores with 1 left for normal everyday use to prevent a slow system or 8 cores for max pointage).


Neodata686: with your computer at 3.6Ghz and the 460 running correctly you should be averaging 26000PPD. Let me know if you need any help. Ill be back tomorrow since I will be on the road in a few.

[Edited on September 21, 2010 at 1:03 PM. Reason : F@H FTW]

9/21/2010 1:01:25 PM

Think i'm picking up a second 460 for SLI because they're so damn cheap. If I had them in SLI mode would SMP just treat it as one? Or should I run two SMP for each?

9/21/2010 1:26:38 PM

SMP is for the cpu only. The gpu3 install is the only thing you need to install for both graphics cards. There is a special setup for SLI but its not hard. FYI...SLI has to be disabled for you to fold on each card individually. Unless they fixed that recently.
1) After you install the gpu3 there will be a directory in C:\Users\"NAME"\AppData\Roaming\Folding@home-gpu that needs to be coppied. AppData is a hidden folder so unhide it. Rename the directories to Folding@home-gpu-0 and Folding@home-gpu-1.
2) Then create two shortcuts to the .exe file for the gpu3 application.
3) Right click, go to properties and change the start in section to be "C:\Users\"NAME"\AppData\Roaming\Folding@home-gpu-0"
4) Add -gpu 0 to the target so it looks like "C:\Program Files (x86)\Folding@home\Folding@home-gpu\Folding@home.exe" -gpu0.
5) Do steps 3&4 for -gpu1.
6) Run the shortcut for -gpu0 and once open go to configure and in the advanced tab change the machine id to 8 that way when you run -gpu1 they wont grab the same id. Just make sure the three clients (1 smp & 2 gpu) dont use the same id numbers and your set.

OK so those instructions were a little longer than intended but they are still manageable.

[Edited on September 21, 2010 at 9:21 PM. Reason : ]

9/21/2010 9:18:54 PM

Wish I could edit after 30min .
BTW there is at least one guy at http://www.evga.com thats folding on an overclocked gtx460 getting 14000ppd each. Though the average for most is closer to 10000ppd. Should be either way. http://www.evga.com/forums/tm.aspx?m=387168&mpage=3

9/21/2010 10:02:43 PM

my graphics card can't use gpu folding. But I did start smp folding on a laptop that I thought was already folding but wasn't.

So that's two machines doing smp folding. Now I gotta get the 4 servers mentioned above folding up more points and it'll be all good.

also, when I get my PS3 back from Sony I'll have it start folding again. That'll help some.

[Edited on September 22, 2010 at 7:15 AM. Reason : holler.]

9/22/2010 7:12:45 AM

If you have any Nvidia card on this list your can fold on the gpu2 or gpu3 client. http://www.nvidia.com/object/cuda_gpus.html

If you have an ATI card that is an x1900 or better you can fold on that as well. I know the newer 5000 series use the flag -forcegpu ati_r700 in the additional client parameters under the advanced tab. There might be a flag for the older cards but I dont know it.

9/22/2010 9:04:51 AM

nope, not on that list. [VIDEO CARD FAIL]

9/22/2010 5:22:55 PM

so wait is it any coincidence that the folding@home team is heating back up now that we're headed into the colder months?

9/22/2010 5:50:33 PM

no comment...

[Edited on September 22, 2010 at 6:10 PM. Reason : ninja rap!!!]

9/22/2010 6:10:05 PM

so here's how I did it on the previously stated machine (the post where I showed the output of cat /proc/cpuinfo

1. First I got the CPU and core info. Decided to use the uniclient instead of the smp client due to the single core output.

[root@han-solo folding]# cat /proc/cpuinfo | grep processor && cat /proc/cpuinfo | grep cores
processor       : 0
processor       : 1
processor       : 2
processor       : 3
cpu cores       : 1
cpu cores       : 1
cpu cores       : 1
cpu cores       : 1

[root@han-solo folding]# mkdir fah0 fah1 fah2 fah3
[root@han-solo folding]# ls
fah0  fah1  fah2  fah3  FAH6.02-Linux.tgz
[root@han-solo folding]# tar -xvf FAH6.02-Linux.tgz --directory=fah0/ && tar -xvf FAH6.02-Linux.tgz --directory=fah1/ && tar -xvf FAH6.02-Linux.tgz --directory=fah2/ && tar -xvf FAH6.02-Linux.tgz --directory=fah3
[root@han-solo folding]# ls *
FAH6.02-Linux.tgz  fahservice  startfolding

fah0:
fah6
mpiexec

fah1:
fah6
mpiexec

fah2:
fah6
mpiexec

fah3:
fah6
mpiexec

[root@han-solo folding]# 

[root@han-solo folding]# chmod +x fah*/fah6 
[root@han-solo folding]# touch startfolding
[root@han-solo folding]# echo "cd /root/folding/fah0/" > startfolding
[root@han-solo folding]# echo "./fah6 -verbosity 9  &" >> startfolding
[root@han-solo folding]# echo "cd /root/folding/fah0/" >> startfolding
[root@han-solo folding]# echo "./fah6 -verbosity 9  &" >> startfolding
[root@han-solo folding]# echo "cd /root/folding/fah0/" >> startfolding
[root@han-solo folding]# echo "./fah6 -verbosity 9  &" >> startfolding
[root@han-solo folding]# echo "cd /root/folding/fah0/" >> startfolding
[root@han-solo folding]# echo "./fah6 -verbosity 9  &" >> startfolding
[root@han-solo folding]# cat startfolding 
cd /root/folding/fah0/
./fah6 -verbosity 9  &
cd /root/folding/fah1/
./fah6 -verbosity 9  &
cd /root/folding/fah2/
./fah6 -verbosity 9  &
cd /root/folding/fah3/
./fah6 -verbosity 9  &
[root@han-solo folding]# chmod +x startfolding 
[root@han-solo folding]# echo "# chkconfig: 345 93 14" > ./fahservice
[root@han-solo folding]# echo "# description: will start FAH as service" >> ./fahservice
[root@han-solo folding]# echo "su - $USER -c \"cd $PWD; ./startfolding < /dev/null > /dev/null 2>&1 &\"" >> ./fahservice
[root@han-solo folding]# cat fahservice 
# chkconfig: 345 93 14
# description: will start FAH as service
su - root -c "cd /root/folding; ./startfolding < /dev/null > /dev/null 2>&1 &"
[root@han-solo folding]# cat startfolding 
cd /root/folding/fah0/
./fah6 -verbosity 9  &
cd /root/folding/fah1/
./fah6 -verbosity 9  &
cd /root/folding/fah2/
./fah6 -verbosity 9  &
cd /root/folding/fah3/
./fah6 -verbosity 9  &
[root@han-solo folding]# chmod +x fahservice
[root@han-solo folding]# echo "/root/folding = $PWD"
/root/folding = /root/folding
[root@han-solo folding]# su - -c "cp /root/folding/fahservice /etc/init.d/"
[root@han-solo folding]# chmod +x /etc/init.d/fahservice 
[root@han-solo folding]# su - -c "chkconfig --add fahservice"
[root@han-solo folding]# 

[root@han-solo folding]# cat fah0/client.cfg 
[settings]
username=RaidenNCSU
team=59
passkey=
asknet=no
machineid=1

[http]
active=no
host=localhost
port=8080

[root@han-solo folding]# cat fah1/client.cfg 
[settings]
username=RaidenNCSU
team=59
passkey=
asknet=no
machineid=2

[http]
active=no
host=localhost
port=8080

[root@han-solo folding]# cat fah2/client.cfg 
[settings]
username=RaidenNCSU
team=59
passkey=
asknet=no
machineid=3

[http]
active=no
host=localhost
port=8080

[root@han-solo folding]# cat fah3/client.cfg 
[settings]
username=RaidenNCSU
team=59
passkey=
asknet=no
machineid=4

[http]
active=no
host=localhost
port=8080
[root@han-solo folding]#

9/24/2010 2:15:50 PM

Im guessing between 800-1000ppd.

9/24/2010 2:43:34 PM

well that's not incredibly good.

correction, echo statements should be:

[root@han-solo folding]# echo "cd /root/folding/fah0/" > startfolding
[root@han-solo folding]# echo "./fah6 -verbosity 9  &" >> startfolding
[root@han-solo folding]# echo "cd /root/folding/fah1/" >> startfolding
[root@han-solo folding]# echo "./fah6 -verbosity 9  &" >> startfolding
[root@han-solo folding]# echo "cd /root/folding/fah2/" >> startfolding
[root@han-solo folding]# echo "./fah6 -verbosity 9  &" >> startfolding
[root@han-solo folding]# echo "cd /root/folding/fah3/" >> startfolding
[root@han-solo folding]# echo "./fah6 -verbosity 9  &" >> startfolding

9/24/2010 2:57:04 PM

I've got a 2x 6-core AMD Opteron I'm trying it on now, and I've taken a different approach. I've got one smp client running per two physical cores, so that's six smp clients on one box. Yesterday's run netted about 2k points, but that's for about a day and a half worth of work. I'm still running big WU's on all six instances since the box has 64 GB of ram.

Anyone know if this can be improved? Say 12 single clients, 3 SMP clients, or just running normal workloads? I've never gotten to play with this on such powerful hardware, so I have no clue how to optimize it.

9/25/2010 8:43:46 AM

The way the bonus system works with the big work units, the earlier you finish processing them the higher the bonus points. There fore you need to give as much processing time to the work unit as possible. One SMP with 12 dedicated cores. Hell I am on an i7 with 4 cores and 4 hyperthreaded cores and it is pulling 15000ppd....and thats not even the -bigadv work units. Just the normal A3 work unit.

9/26/2010 3:33:20 AM

^^I'd post that one here: http://foldingforum.org/


and whoever posts the mega post again, update the F@H Forums URL

9/26/2010 10:37:15 AM