• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

SMP PPD on C2D vs X2?

M

Mohonri

Supreme [H]ardness
2FA
Joined
Jul 29, 2005
Messages
5,774
I'm going to build a new machine in the near future, and at least for the interim will be installing Linux and the SMP client. I've heard that the SMP client for C2D will net me around 1000ppd. What about on an X2?
 
search for [BRO]Alaskan's posts in here. He has a bunch of AM2 chips running the SMP client. His PPD is rather impressive at ~4K.
 
I get ~950PPD from both my 939pin and AM2 3800's O/C to 2.65ghz. Both have single Ch 512 meg Ram.
I get ~860 PPD from both of my AM2 4200's O/C to 2.4ghz (cheap boards with on board video) and 512 meg single ch ram.
I have seen posts from C2D's running SMP getting 1300-1500PPD
 
My lappy with a C2D 7200 (2ghz) is pounding my x2 4600+ (2.4ghz) in the SMP client.

That is all.
 
My E6400@3.2 would put out 1500+ PPD on the SMP client if I would stay in Linux. For now it's a bit lower since I still use XP for gaming and a few other things. I just haven't gotten around to messing with Wine just yet although I need to.

 
What about running the SMP client in VMWare, and adding a GPU client running in Windows? Is that permissible? Does it give a significant boost?
 
I get ~750 PpD from dual Opterons @2.4Ghz.
I get ~2,000 PpD from my dual dual Opterons @2.6 Ghz.

If a C2D system can put out silly numbers (+1,000PpD) running the SMP client I may have to swop my boxen's clients around.

Luck ........ :D
 
Tigerbiten said:
I get ~750 PpD from dual Opterons @2.4Ghz.
I get ~2,000 PpD from my dual dual Opterons @2.6 Ghz.

If a C2D system can put out silly numbers (+1,000PpD) running the SMP client I may have to swop my boxen's clients around.

Luck ........ :D
Click to expand...

look at my december stats: me folding full force
look at my january stats: me folding full force + 1 C2D @ 2.8 VMWare SMP.
 
Tigerbiten said:
I get ~750 PpD from dual Opterons @2.4Ghz.
I get ~2,000 PpD from my dual dual Opterons @2.6 Ghz.

If a C2D system can put out silly numbers (+1,000PpD) running the SMP client I may have to swop my boxen's clients around.

Luck ........ :D
Click to expand...

C2D are doing better clock per clock and can O/C better. I can see a single quad core easly getting 2800PPD from this build

I'm planning on a (hopefully) 2800PPD build when the QX6600's drop in price.
QX6600...............~$500
GA-965P-S3........~$111
1gig DDR2-800......$105
HD I have an 8gig to use
case.....................$25
Powersupply I have a 450w to use
ATI X600...............$20

Total ~$766 my goal is <$800 and if I have to go aftermarket on the cooler I'll be close
I plan on O/Cing (of course) to ~3.0ghz
 
Mohonri said:
What about running the SMP client in VMWare, and adding a GPU client running in Windows? Is that permissible? Does it give a significant boost?
Click to expand...

Yep. I'm doing that with my primary Windows XP machine. It's a C2D 6700 @ 3.4 GHz with a X1950XTX running the SMP client in VMWare along with the GPU client.

It varies, but in general that combo gives ~1800-2000 ppd.

VMWare only leaves about 20% CPU left for the GPU which hinders its performance
by ~30%, but it's still the best ppd option for a dual core cpu + high end GPU
running Windows.

I also have 2 dedicated C2D SMP Linux (SUSE 10.2) machines and they
typically crank out 1200-1400 ppd.

For < $400 you can build a dedicated C2D SMP machine that will produce well over
1000 ppd.
 
mklinger said:
I also have 2 dedicated C2D SMP Linux (SUSE 10.2) machines and they
typically crank out 1200-1400 ppd.

For < $400 you can build a dedicated C2D SMP machine that will produce well over
1000 ppd.
Click to expand...

OK, I just built a dedicated E6400 SMP machine.... struggling a bit with installation , but getting there.... question is:

Do I install FAH as root , or will it work installed as a user ? Anything in particular I need to do regarding permisions if I install it as a user ?

Thanks, RPHArrow

 
Well my 2.0Ghz Core Duo Macbook gets around 704 PPD. Since we know that C2D is faster than the Core Duo, and you can overclock, you will certainly easily get to 1K PPD.
 
RPhArrow said:
OK, I just built a dedicated E6400 SMP machine.... struggling a bit with installation , but getting there.... question is:

Do I install FAH as root , or will it work installed as a user ? Anything in particular I need to do regarding permisions if I install it as a user ?

Thanks, RPHArrow

Click to expand...

I suggest that you use the finstall script:
as user:
Code: 
wget http://ranume.radionet.ee/folding/finstall 
chmod +x finstall
./finstall smp
...wait for it to finish
Code: 
sudo ./installservice

This will make it autorun, as limited user.
 
mklinger said:
Yep. I'm doing that with my primary Windows XP machine. It's a C2D 6700 @ 3.4 GHz with a X1950XTX running the SMP client in VMWare along with the GPU client.

It varies, but in general that combo gives ~1800-2000 ppd.

VMWare only leaves about 20% CPU left for the GPU which hinders its performance
by ~30%, but it's still the best ppd option for a dual core cpu + high end GPU
running Windows.
Click to expand...

Question: How long is it taking you to get through an SMP WU on average and a GPU WU on average? The GPU has a perfered return of 3 days, and the SMP anywhere from .75 to 2 days.
 
ICE_9 said:
Question: How long is it taking you to get through an SMP WU on average and a GPU WU on average? The GPU has a perfered return of 3 days, and the SMP anywhere from .75 to 2 days.
Click to expand...

It really depends on the SMP WU. For the smaller (587 and 716 pt) ones, the SMP core runs about 5:30 to 7 min/frame and the GPU runs at ~8 min/frame at the same time.

For this new monster 1760 pt SMP WU, if I run them together, the SMP frame times are ~ 12:30 which is pretty good, but the GPU bogs down to almost 30 min/frame!

If I know I've got a 1760 pt unit, I'll stop the GPU and get the SMP frame times to ~11:30.

On this machine, if I run the GPU client only and dedicate a full CPU core, I'll see frame times around 5:30.
 
RPhArrow said:
OK, I just built a dedicated E6400 SMP machine.... struggling a bit with installation , but getting there.... question is:

Do I install FAH as root , or will it work installed as a user ? Anything in particular I need to do regarding permisions if I install it as a user ?

Thanks, RPHArrow

Click to expand...


I've got mine running as root, but it really shouldn't matter. If you set up user accounts, you should be able to fold under those accounts without any problem.
 
aside:
My 2.25GHz Opty 165 is currently taking ~9 mins/ frame on an SMP WU. Debian Etch x64.
 
My Opty 240's are taking 9:45 a frame & my Opty 285's are taking 10:30 a frame.

But my 240's are crunching a p3025 (652 points) and my 285's are crunching p3036 (1440 points) hence the relative time per frame difference.

Luck ......... :D
 
You must log in or register to reply here.
Back
Top