I'm legal now! Fold on

tank1023

I'm legal now! Fold on

 
congrats to you too
 

just noticed what I hope is a cool feature regarding the forced download of 2.25.   I replaced 2.25 with 2.22 in both Fermi core folders, Beta and Core_15.fah.  I received a wu that forced 2.25 back into V7.  When I tried to rollback to 2.22 again I noticed that the Beta folder still has 2.22 and the Core_15.fah folder has 2.25.  I'm currently working 8018 and 8057 and both are using 2.22. The only wu that used 2.25 was the one that forced the core. I hope that makes sense, basically you don't have to worry about 2.25 being forced, 2.22 will (should) stay in the beta folder. Keep in mind, this is for Fermi only.

widsss

just noticed what I hope is a cool feature regarding the forced download of 2.25.   I replaced 2.25 with 2.22 in both Fermi core folders, Beta and Core_15.fah.  I received a wu that forced 2.25 back into V7.  When I tried to rollback to 2.22 again I noticed that the Beta folder still has 2.22 and the Core_15.fah folder has 2.25.  I'm currently working 8018 and 8057 and both are using 2.22. The only wu that used 2.25 was the one that forced the core. I hope that makes sense, basically you don't have to worry about 2.25 being forced, 2.22 will (should) stay in the beta folder. Keep in mind, this is for Fermi only.

 
This would be good info to pass to the beta team (if they don't already know it of course...and if they did they should shared that info ).
 
I may have missed it all the posts, but what advantage does 2.25 bring? I though it was a step toward moving toward optimizing Kepler, but it appears that it doesn't like Fermi as a result. Since the client is able to tell what kind of GPU it's running, why can't they just have separate folders with the appropriate core? Much like widss has now. That way it gets the best preduction out of the equipment.
 
Are these new wu's being written with the 2.25 core in mind? (that would be my guess). Are there any ill effects by forcing it to use 2.22?

kody7839

This would be good info to pass to the beta team (if they don't already know it of course...and if they did they should shared that info ).

I may have missed it all the posts, but what advantage does 2.25 bring? I though it was a step toward moving toward optimizing Kepler, but it appears that it doesn't like Fermi as a result. Since the client is able to tell what kind of GPU it's running, why can't they just have separate folders with the appropriate core? Much like widss has now. That way it gets the best preduction out of the equipment.

Are these new wu's being written with the 2.25 core in mind? (that would be my guess). Are there any ill effects by forcing it to use 2.22?

 
I'm not sure why only one folder was updated with 2.25. I have 2 gpus running, perhaps the other 2.22 was being used at the time, so only one was replaced. All I have are baseless theories, until someone who knows what they're talking about informs me of the real answer. My hope is that the Beta folder is left alone during the core update, but that'd be too good to be true.

GTX480 shader clock 1470
Avg 2:01 133.2k ppd
Min 2:00 134.8k
Final credit estimate, 18451.00 points

Just received an 8018 but nevermind! Can't have good units everytime!

I'm just getting no love here..
 
4 GPUs all running P7623 & P7623 (Half PPD WUs). Haven't seen a single 8057. Both Bigadv rigs are averaging under one 8102 WU per month..

I had a 7623 a few days ago. Not only do you take a ppd hit, but it also updates your core folder.

the qrb discussion thread seems to be missing
 
Anyone from Beta team can pass along the following info
660ti @ 1097/3055
 
 Project ID: 8057
 Core: OPENMMGPU
 Credit: 2549
 Frames: 100


 Name: local Slot 00
 Path: 127.0.0.1-36330
 Number of Frames Observed: 261

 Min. Time / Frame : 00:02:17 - 110,558.1 PPD
 Avg. Time / Frame : 00:02:17 - 110,558.1 PPD
 Cur. Time / Frame : 00:02:17 - 110,558.1 PPD
 R3F. Time / Frame : 00:02:17 - 110,558.1 PPD
 All  Time / Frame : 00:02:17 - 110,558.1 PPD
 Eff. Time / Frame : 00:02:19 - 108,180.5 PPD
 
comparison data (same card)
 Project ID: 8018
 Core: OPENMMGPU
 Credit: 5757
 Frames: 100


 Name: Tracker 660ti - core 2.22
 Number of Frames Observed: 143

 Min. Time / Frame : 00:04:11 - 19,816.9 PPD
 Avg. Time / Frame : 00:04:25 - 18,770.0 PPD

Just to add in case anyone wants to know. 
GTX 460 768MB @ 850 Core
WU 8057
Core_15 v2.25
PPD 72xxx  3 min 20 sec per fold
Core_15 v2.22
PPD 107xxx 2 min 22 sec per fold
 
over 30k PPD gain going back to v2.22 and shaving off almost 1 min per fold is just insane. 

I get a better tpf if I'm folding two 8057s at the same time, 1.23 tpf. When I'm folding a 8018 and 8057, the QRB Wu jumps between 95-97%,with a 1.27 tpf.

tank1023

hjohnson....how are you getting such a low TPF on a 550?
My 550 is like 5 min. I'm also running a SMP and two GPU's on this rig.

Either your still on 2.25 core or you have other problems?
Or maybe he meant 560ti?
Edit: This is from my GTX460m laptop video card:
Running -smp too. times are actually lower while smp folding for the GPU. avg 3:52
My idle clock for the CPU is 800Mhz & have I it locked to 2.0Ghz under load to keep it under 80c on the CPU.
So the CPU clock is playing a part to how the GPU times are at least for this rig. GPU avg is 3:57 with CPU idle.
 
 Project ID: 8057
 Core: OPENMMGPU
 Credit: 2549
 Frames: 100

 Name: v7-CPU8-460m Slot 00
 Number of Frames Observed: 300

 Min. Time / Frame : 00:03:51 - 50,495.6 PPD
 Avg. Time / Frame : 00:03:54 - 49,527.6 PPD

No idea. Yes, this is the 550 Ti, not the nicer 560 or 560 Ti. It works fine for what I use it for but is not great for folding. Why mine might possibly be faster than its peers, a few wildly uneducated and speculative guesses that will be totally wrong:

Because it's a 550 Ti. Being made by Texas Instruments makes it way faster.
Running multiple GPUs may slow things down.
I may be avoiding some memory bottleneck by having the larger 2GB card. (seriously doubt it)
Probably because I'm using Windows 7 64 bit/F@H v7.1.52/driver version of thing 2.7.whatever, which is the fastest version. (No.)
Aggressive fan curve set up with EVGA Precision to avoid thermal throttling? No, I didn't think so either.
I did the reinstall core 2.22 trick after my first beta units forced 2.25 on me. There is definitely a difference between 2.22 and 2.25. I do not believe there is a difference between using 2.22 in beta and not, as getting 2:58 per frame WUs is common for me. No idea if those were 8057s but they were ~2,300 point WUs.
 
So, my guess is try running it on a single card and see what happens. My setup is just an i3-2320 and the 550Ti, and while running the GPU does slow down the CPU TPF slightly, the reverse does not seem to be the case.
 

hjohnson1

Because it's a 550 Ti. Being made by Texas Instruments makes it way faster.

 
That just made my day, TY sir