In regards to the new v6.30 drop-in binary...

augie1111 · Jul 25, 2010

Any comments from the peanut gallery on this? :blah:

I haven't loaded it yet though it seems to only be a marginal help in setup with the main feature being the ability to run in services, which I really like, set it and forget it especially for my Q6600 which is now dedicated to folding.

ChasR mentioned in his guide to use -advmethods, does this mean I can't run -bigadv then using this drop-in in my prepped i7 to run as a service?

Thanks as always good folk! :beer:

harlam357 · Jul 25, 2010

Yeah, you can run -bigadv no problem... it's just that that users without i7s (or other 8-core systems) will just need to use -advmethods.

augie1111 · Jul 25, 2010

harlam357 said:
Yeah, you can run -bigadv no problem... it's just that that users without i7s (or other 8-core systems) will just need to use -advmethods.

Thanks harlam, this is as I assumed it was.

Just making sure as I'm on the last legs of the current -bigadv, don't want to go and lose that eh?

ChasR · Jul 25, 2010

Actually, you don't have to use -advmethods anymore. When there were a1 WUs, you had to use it to get a3 WUs. Now there are only a3 WUs and the flag is unnecessary.

I'll clean that up in a bit.

I'm not sure it's worth the 2000 ppd gain to fold -bigadv on an i7. It comes so close to the preferred deadlines that if anything happens that delays the return, you loose a lot of points.

augie1111 · Jul 26, 2010

ChasR said:
I'm not sure it's worth the 2000 ppd gain to fold -bigadv on an i7. It comes so close to the preferred deadlines that if anything happens that delays the return, you loose a lot of points.

I hear ya loud and clear. Due to BSODs, power outs and ISP issues, my last WU just made the deadline with 11 minutes to spare! :bang head

Lancelot · Jul 26, 2010

I noticed that after closing the new 6.30 client, it always restarts with "Previous termination of core was improper. Going to use standard loops."
I'm closing the client using control-c

Also I uninstalled the MPICH driver/program under software, but still a Windows firewall dialog comes up upon starting the 6.30 client pointing to system32. How do I kill & clean this thing?!

augie1111 · Jul 26, 2010

Another -bigadv anecdote, I posted it here as it's not an issue anymore. Got a new 2684 yesterday and all looked fine with the TPF @ 44min. but then it just kept degrading from there until it ended up @ 57m TPFtoday which put me way past the deadline. Maybe a bad WU, IDK but I'm not a fan of it anymore.

Oh, I got the shock of the day when I first installed v6.30 today, gave me an a0 to chomp on. What?! :fight:

Well, the parameters were correct upon a doublecheck so I just deleted the proper files and got my regular 6012 back. Weird. :shrug:

ChasR · Jul 26, 2010

All the various SMP clients will return improper termination of the core messages. Fact is SMP won't run without optimizations. The message is a cosmetic error. I went through the registry and removed all MPICH keys. I still get the errors reported in the console on startup but they have no effect on operation.

B2K24 · Jul 26, 2010

ChasR said:
I'm not sure it's worth the 2000 ppd gain to fold -bigadv on an i7.

Considering bigadv consists of P2684 and P2685 I believe it's totally worth it if the system is 24/7 and dedicated folder.

The P2685 have proved to be AWESOME and I will take my chances with that over bad 6701 and 6702.

Audioaficionado · Aug 13, 2010

I've had none of the problems with v6.30 binary drop in I used to have with the earlier MPICH required versions. They all caused me at least one BSoD per day.

Too bad -bigadv isn't practical on my Q6600@3GHz system.

sfu7274456 · Aug 13, 2010

ChasR said:
Actually, you don't have to use -advmethods anymore. When there were a1 WUs, you had to use it to get a3 WUs. Now there are only a3 WUs and the flag is unnecessary.

I'll clean that up in a bit.

I'm not sure it's worth the 2000 ppd gain to fold -bigadv on an i7. It comes so close to the preferred deadlines that if anything happens that delays the return, you loose a lot of points.

What am I missing here. I have heard this before, I get 15K per day on my I7's & 3.6ghz without the bigadv flag, and I get 26K per day on the same machine running bigadv

harlam357 · Aug 14, 2010

What project are you running sfu?

ChasR · Aug 14, 2010

sfu7274456 said:
What am I missing here. I have heard this before, I get 15K per day on my I7's & 3.6ghz without the bigadv flag, and I get 26K per day on the same machine running bigadv

On the machine I tested on a3 bigadv got 20,000 ppd while regular a3 got 18,000 ppd. I was using an i7 @ 3.8 ghz. I don't have that machine any more so I can't say how it does on newer a3 bigadv WUs.

sfu7274456 · Aug 14, 2010

On the 2684's I get basically them same PPD's as regular WU's. But it seems to me like their are as not as many 2684's out there anymore. Of my 8 machines currently running only one is running 2684's, the rest are getting 2685's, and 2686's and my ppd are running between 22K and 28K per day and all are running right around 3.6Ghz. Maybe I just answered my own question. I think what I was getting at is it sounded to me like people were getting around 20K per day on regular a3's.

In regards to the new v6.30 drop-in binary...

augie1111

Member

harlam357

Senior Fold-a-holic

augie1111

Member

ChasR

Senior Member

augie1111

Member

Lancelot

Member

augie1111

Member

ChasR

Senior Member

B2K24

New Member

Audioaficionado

Sparkomatic Moderator

sfu7274456

Member

harlam357

Senior Fold-a-holic

ChasR

Senior Member

sfu7274456

Member

Similar threads