F@H - my 1000th WU

Associate
Joined
27 Apr 2010
Posts
100
is there a special ritual or custom that you guys do when you hit your 1000th Folding@home work unit?

i was somewhere up at 990 work units when i took a break from folding over a year ago, (was using a netbook and the heat on my palms was unbearable) and now that i have the gpu client its been munching through work units like real fast , like....

so after a day and a half of gpu folding, im already at my 998th WU, on the old SMP client i was using on my q6600 this number would have taken a month and a half to reach
 
thanks

not quite at the big 1000 yet tho, one and a half work units left to go, which is roughly half a day more
i would like to start planning for the occasion so that i will be ready when the 1000th unit is submitted
 
is there a special ritual or custom that you guys do when you hit your 1000th Folding@home work unit?

Yes. I believe you go on to do your next Folding@home work unit, and the next one after that, and the next one after that, and...

Congrats on the milestone, 'tis a mighty fine effort and no mistake :)
 
Well, there's always a certificate that you can get from Stanford to mark the occasion that you can proudly flaunt around here, but other than that I'm afraid you just plug away as usual, issuing parps to - and then stomping - slackers into submission as you go (which is fun in itself, of course :D).
 
To 'parp' someone is warning them that you'll be passing them in the rankings in the near future. It's generally used to gee up active members in the hope that they'll up their production, although usually they just continue with their slack ways and let you sail right past them.

That said, feel free to parp all and sundry :)
 
Yep, it counts!

I wouldn't say it's unfair at all. It's certainly not cheating and it was your decision and dedication to use your GPU and spend the extra leccy costs in running it. Besides, it's not your fault if those that you're about to stomp are too slack to go down the GPU route themselves, so go for it :p
 
ohh snap! i forgot to check the meter today... dont want the room to get dark all of a sudden

thing is, the GPU spends so little time to finish a WU , while the CPU clients steam away at 100% load for weeks on end, so i feel that the electricity cost : WU completed ratio is heavily in the favour of GPU folding...
 
Last edited:
While GPU folding is better for mass producing points an OC'd i7 920/930 will still produce the same PPD on A3 units as a GTX480, even more if you bigadv on the CPU
 
cert.php



i see that the SMP client isnt using more than 13% cpu, does anyone know why that is?
 
The GPUs do more WUs because the WUs are smaller on the GPU clients - simple as. Points is what really counts :p

What CPU? If it's an i7 then it might only be using one thread.
 
i see that the SMP client isnt using more than 13% cpu, does anyone know why that is?

I suspect it's because you're not running it with the -smp flag.

You need to create a shortcut to the "Folding@home-Win32-x86" executable, right click the shortcut and select properties and add -smp to the end of the "Target" field like I've done below. You then need to use this shortcut to start the client in order to run SMP units which use all of the CPU and give much better ppd.

Shortcut.JPG


Well done on 1,000 WUs btw :)
 
While GPU folding is better for mass producing points an OC'd i7 920/930 will still produce the same PPD on A3 units as a GTX480, even more if you bigadv on the CPU

All I need is a mobo which will overclock that can take 4 hyperthreading hex core Xeons and costs less than £1500.00 :D
 
I suspect it's because you're not running it with the -smp flag.

You need to create a shortcut to the "Folding@home-Win32-x86" executable, right click the shortcut and select properties and add -smp to the end of the "Target" field like I've done below. You then need to use this shortcut to start the client in order to run SMP units which use all of the CPU and give much better ppd.

ahh thanks... turns out that so far i was using fahcore_78 , after ive set up a new shortcut with the -smp flag, it said that the 78 core was no valid for this type of work or something along those lines... so it downloaded fahcore_a0, then it got some core io error about the size of the work units or something, i had to shut down mpiexec.exe, delete the 'work' folder, and launch F@H again... this time it download fahcore_a3, as well as a new work unit

now i have cpu utilization at 100%, but i dont see the 4 instances of fahcore_ax that the F@H guide says, just one instance taking up ~80% cpu while the other 13% is the gpu folding client

hmm mpiexec.exe is not running, i hope that wouldnt be a problem... gonna restart my computer just in case

okay just restarted, still no mpiexec.exe ... the clients are still taking up 100% cpu but i dont want to come back this evening to find that some error happened because mpiexec.exe wasnt running...
 
Last edited:
Back
Top Bottom