Looks like it may be time to switch to v7 of F@H

Could be a goldfish.

This morning I cleaned the dust out of the rad and reapplied TIM. Unfortunately I did not find any goldfish. :(

Now that the loop is all tweaked it's time give v7 another go with the 980x.
 
I saw this post and decided to make the switch to v7, but man the 7610/7611 WU is depressing. Halved my PPD and pushed my temps up almost 10C.
 
Same here. After upgrading(?) to v7, my i7's performance has plummeted, with respect to PPD. It seems as though the A4 core projects include everything under the sun, even dual-purpose uniprocessor/SMP units, such as 10085. A4 appears to be a catch-all, general purpose core. I am not impressed.
 
What flags did you set for the client and the core? Also, did you set the core count to the actual number of threads or did you leave it at default of "-1"?
 
THat's odd, most of the A4 WU's I get seem to make tons of PPD.

If you have a hyper threaded 6-core, sure...if you have anything less, then even the crappy AMD GPU client can cook off better ppd than smp.
 
I'm considering uninstalling V7 and returning to V6.34 SMP. My I7 system is down about 30% in production.
 
If you have a hyper threaded 6-core, sure...if you have anything less, then even the crappy AMD GPU client can cook off better ppd than smp.

Well, my i7 980 does bigadv, and the Sandy-E is off for the summer, but several other boxes from dual cores to quads all seem to do really well with A4 core WU's.
 
I7 870 @ 3.8GHz
Win 7/64
SMP
parameters: advanced, SMP 8, verbosity 9

apart from the thread count (SMP X), are you guys using different flag assignments?
 
I was really trying to like v7, but after multiple core crashes I called it quits and went back to 6.34. :mad:
 
I just wanted to post up some data to expand on my comment. This is on an 2600k at 4.9Ghz

Project ID: 7610
Core: GRO-A4
Credit: 788
Frames: 100
Number of Frames Observed: 300
Min. Time / Frame : 00:03:17 - 27,692.9 PPD
Avg. Time / Frame : 00:03:56 - 21,120.3 PPD

Project ID: 7611
Core: GRO-A4
Credit: 788
Frames: 100
Number of Frames Observed: 200
Min. Time / Frame : 00:04:48 - 15,666.7 PPD
Avg. Time / Frame : 00:05:20 - 13,376.5 PPD
 
For comparison, WUs which I completed with V7/a4 core. i5 [email protected] GHz. (VM/Ubuntu 12.04/64b; without GPU):

p7000 - 7028 - 39-46k PPD
p7083 - 7084 - 15.5k PPD
p7610 - 18.5k PPD
p7611 - 9.3k PPD
p8001 - 8004 - 17k PPD
p8041 - 8042 - 23k PPD
p8046 - 8048 - 23k PPD
p10062 - 10090 - 41k PPD

This processor with V6.34 gives ~ 21k PPD.
The problem is not in V7, but in the basic points of WUs with a4. These WUs with 40k PPD come very rarely, and not compensate the losses from p7611. p7611 server assigns it often.
I returned yesterday to V6.34. +10% does not compensate the losses.

Sorry for bad English
 
For comparison, WUs which I completed with V7/a4 core. i5 [email protected] GHz. (VM/Ubuntu 12.04/64b; without GPU):

p7000 - 7028 - 39-46k PPD
p7083 - 7084 - 15.5k PPD
p7610 - 18.5k PPD
p7611 - 9.3k PPD
p8001 - 8004 - 17k PPD
p8041 - 8042 - 23k PPD
p8046 - 8048 - 23k PPD
p10062 - 10090 - 41k PPD

This processor with V6.34 gives ~ 21k PPD.
The problem is not in V7, but in the basic points of WUs with a4. These WUs with 40k PPD come very rarely, and not compensate the losses from p7611. p7611 server assigns it often.
I returned yesterday to V6.34. +10% does not compensate the losses.

Sorry for bad English

I agree, that's why I have not switched to V7, Staying with V6.34. :D
 
Now I see the announcement, which should have been done long ago.

Project 7611 Points Increased

Base points value has been increased from 788 to 1400.

Thank. Gawd.

About damn time. Now that they've raised the base points, the units will probably cease to exist in a few weeks, and be replaced by units in line with the 788 base pts....What is really funny is that it took them so long, and to bring them to parity they had to almost double the base value.
 
Now I see the announcement, which should have been done long ago.

Project 7611 Points Increased

Base points value has been increased from 788 to 1400.

Nice. Since I changed over to v7 2 days ago I've gotten nothing but 7611's and 7610's. The worst thing is that the actual work it's doing isn't really that motivational.

Edit: I just got a 8050 WU- normal points And related to the cause I fold for. Woot
 
Last edited:
For comparison, WUs which I completed with V7/a4 core. i5 [email protected] GHz. (VM/Ubuntu 12.04/64b; without GPU):

p7000 - 7028 - 39-46k PPD
p7083 - 7084 - 15.5k PPD
p7610 - 18.5k PPD
p7611 - 9.3k PPD
p8001 - 8004 - 17k PPD
p8041 - 8042 - 23k PPD
p8046 - 8048 - 23k PPD
p10062 - 10090 - 41k PPD

This processor with V6.34 gives ~ 21k PPD.
The problem is not in V7, but in the basic points of WUs with a4. These WUs with 40k PPD come very rarely, and not compensate the losses from p7611. p7611 server assigns it often.
I returned yesterday to V6.34. +10% does not compensate the losses.

Some kind of coincidence... your posting about p7611 getting such low points, then shortly after TJ changes the points for no apparent reason, and on a Sunday. :D Good feedback results in good improvements.

Are you going to give the V7 client another try?
 
Last edited:
Some kind of coincidence... your posting about p7611 getting such low points, then shortly after TJ changes the points for no apparent reason, and on a Sunday. :D Good feedback results in good improvements.

Are you going to give the V7 client another try?

Yes 7im,
just went back to V7. After 2-3 days I will write here for PPD, with new base points of p7611.
 
7im, I promised to report after a few days. V7 is currently working with long series of p8004, p7000-7028 less, and rarely p7645-7647. Average PPD is about ~ 19k (+10% included; with V6.34 ~ 21k PPD). I do not see p7610/11. VM/Ubuntu OS,"advanced" flag is removed. For now I will continue with the client V7 :)
 
Back
Top