• 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.

bigadv units exists anymore?

T

thegtproject

Weaksauce
Joined
Jun 18, 2011
Messages
126
Do bigadv units exist anymore? I got a 12 core CPU now but seeing low results:

http://folding.extremeoverclocking.com/user_summary.php?s=&u=568349

Using latest Linux FAH client with this config.xml

<config>
<!-- Configuration file created by FAHClient on 2012-05-05T09:21:43Z -->
<user value='thegtproject'/>
<team value='33'/>
<passkey value='13b9b4c4ab12aa336c4c51ebb564b157'/>
<smp value='true'/>
<gpu value='false'/>
<client-type value='advanced'/>
</config>
 
you need to set max-packet-size to big.

I have been getting the occasional p6900 on 2 of my windows v7 clients (i7 3930k, and i7 920)
 
g/l on linux getting BA as I think theyre all 16 core 8101s now, with some 6900 stragglers.
 
For bigadv, you need to set advancedmethods off and use the -bigadv flag

e.g. -smp 12 -bigadv -verbosity 9
 
Biffa said:
For bigadv, you need to set advancedmethods off and use the -bigadv flag

e.g. -smp 12 -bigadv -verbosity 9
Click to expand...

I'm running these on my i7 970, and get 6903/4s about 75% of the time.

linuxforge image, and the Folding client 6.XX that comes with the image.
 
jfb9301 said:
I'm running these on my i7 970, and get 6903/4s about 75% of the time.

linuxforge image, and the Folding client 6.XX that comes with the image.
Click to expand...

I used to get 6903/4's but recently have hit a big string of 6901's on my i7 980 / linux.
 
extide said:
you need to set max-packet-size to big.

I have been getting the occasional p6900 on 2 of my windows v7 clients (i7 3930k, and i7 920)
Click to expand...

Code: 
max-packet-size         <string=normal>         Max size in bytes of a work unit packet. Can be
                                                  small=5MB, normal=10MB, big=500MB or a number.

Worth setting max-packet-size to anything higher? I got 32 GB of memory on this rig, don't know if it'll help or not.
 
402blownstroker said:
Unless you want to limit the number of cores used, just use '-smp'. -bigadv gets the turdballs... I mean 8101 WU ;)
Click to expand...

-bigadv only gets the 8101 if you have more than 16 cores.

thegtproject said:
Code: 
max-packet-size         <string=normal>         Max size in bytes of a work unit packet. Can be
                                                  small=5MB, normal=10MB, big=500MB or a number.

Worth setting max-packet-size to anything higher? I got 32 GB of memory on this rig, don't know if it'll help or not.
Click to expand...

Yup, set it to "big"

thegtproject said:
Just got an 8014, not sure if right or not?
Click to expand...

Well 8014 is a higher than normal return standard SMP WU, I'm pretty sure its not a bigadv WU as all my bigadv only rigs have never crunched one.
 
Last edited:
thegtproject said:
So you're saying to set-packet-size to "big" instead of say... "20GB" as the option does allow for specific values?

Also, a big wtf to http://folding.extremeoverclocking.com/user_summary.php?s=&u=568349... see that 1 WU for 400 pts and 0 for the hours before and after? I have no idea why that happens.
Click to expand...

Are you running GPU? GPU units are not very big and complete quickly. You may have had something go wrong with the client or graphics card.

Set the packet size to big, there are only 3 values for that setting. Having more memory than what is needed for the specific workunit/class of work unit does not provide any more performance. The client will not use more memory if you have more than is required. The number of threads, and specific work unit have more effect on the memory footprint.
 
Untitledone said:
Are you running GPU? GPU units are not very big and complete quickly. You may have had something go wrong with the client or graphics card.

Set the packet size to big, there are only 3 values for that setting. Having more memory than what is needed for the specific workunit/class of work unit does not provide any more performance. The client will not use more memory if you have more than is required. The number of threads, and specific work unit have more effect on the memory footprint.
Click to expand...

Nah, I do not have the GPU even running FAHClient. This is all being done on a 3.8 ghz i7-3930K. I may enable the GPU in the future, if it's worth running a GTX 680 on WUs
 
thegtproject said:
Nah, I do not have the GPU even running FAHClient. This is all being done on a 3.8 ghz i7-3930K. I may enable the GPU in the future, if it's worth running a GTX 680 on WUs
Click to expand...

That points output looks very odd. I am not sure what your problem is at this time. This is all on SMP with the -smp 12 flag?
 
Post the output of your FAHlog.txt that should help us troubleshoot :)
 
What do you guys think about my current PPD hourly production since that little hiccup? Do you think I'm on par for a 3.8 GHz i7-3930K ? If not, I'll start posting logs and we can debug. I want to contribute as much as I can :)
 
Yes, that output looks much better. Give it some time and it will even out if you leave it folding.
 
You must log in or register to reply here.
Back
Top