How the heck does that happen? Bad enough I've been getting some 6901's lately, but that's ridiculous. Anyone else experiencing the same? Has to be a temp work server outage.

As always the server will send out SMP work if it has no -bigadv work available, you must have caught it at the wrong moment.  I pulled a 6903 off teher a couple of hours ago - but like you i have had a fair few 6901's in the last few days

I did get one a few days ago but it was dispatched by the rig it was on in less than 12 hours so no harm no foul.

I haven't seen one at all.. I'm using "-smp 48 -bigbeta" setup in the config with Ubuntu 12.04.

Well, two more 6901's followed now by a 8101. Some [H] members seeing the same thing. So much for "-bigbeta". The only place i'm getting 6903 and 6904 consistently is on my 970. Time will tell if this is a trend or just a short term shortage. Figures. Same thing happened when I built more quad I7's to get in on -bigadv. Turned out I was coming in when the party was just wrapping up. Always behind the curve it seems when it comes to folding.

I also picked up an 8108 last night. This sucks.
I feel the same way, always showing up when the parties over.

This is really bad, now all I'm getting are 8108's. I know it's about the science but my PPD is taking a nose dive.

The only consolation is knowing that everyone is faced with the same thing. That includes other teams. At least the PPD dive will be proportional across the board.

mflanaga


The only consolation is knowing that everyone is faced with the same thing. That includes other teams. At least the PPD dive will be proportional across the board.

If the team percentages of points from bigadv were the same, that would be true, but I think certain other teams have a much higher portion of their points coming from bigadv, and will have a bigger drop.  Not that it feels any better from an individual perspective...

Punchy

mflanaga


The only consolation is knowing that everyone is faced with the same thing. That includes other teams. At least the PPD dive will be proportional across the board.

If the team percentages of points from bigadv were the same, that would be true, but I think certain other teams have a much higher portion of their points coming from bigadv, and will have a bigger drop.  Not that it feels any better from an individual perspective...

I meant it on an individual basis within the teams "across the board", but certainly agree with your point, Punchy. 

Gotcha, sorry for the misinterpretation.

Good point, if we are getting 8108's then so is everyone else.

Horvat


Good point, if we are getting 8108's then so is everyone else.

...and it seems to be hurting [H] more than it is us...
http://folding.extremeoverclocking.com/team_summary.php?s=&t=111065

http://folding.extremeoverclocking.com/team_summary.php?s=&t=33

Unfortunately I have all 8108's now except my slowest rig. My PPD is almost half of what it was. The other bad news is when that last rig picks up a 8108, I will have to shut it down or convert to regular -smp as it cannot finish the 8108's on time. Anyone have have a pair of X5660's or better they can donate for a good cause?

My 875k is picking up some crap smp's that have a TPF of 20 min. These are taking a day to finish and only giving me around 15k on that rig. I may have to fire up the GPU's
BUT... It's for science not points so I'll just ride it out.

Yup.

With linux you can move a good wu from a slow machine to a faster one, or finish up a wu from a slow machine that won't quite make it on a faster machine to keep the bonus.
 
Unlike windows where moving WU's is a nightmare, with linux you move the entire folder, user ID and all,  from one machine to the other.  As the user ID in Linux is assigned to the folding folder rather than to the computer Registry, like it is in windows.
 
A couple of things you need to look out for. 
 
1.  Before you move a wu, make several copies of your Folding folder at different time in quick succession.   Some times they don't start up right so I usually make 3 copies.  Don't stop the client, just copy and past while it is folding.  (I have never lost a wu following all these steps)
 
2.  Before you start any work unit you have moved (I do any time I re-start a wu) make sure you disconnect the Internet.  That way if it doesn't work, and some won't, you haven't let it be reported as a failed wu to Stanford, so it gets reassigned and you get no credit.
 
3.  Make sure you have the same plug ins installed on all machines, like the Krakin as even if you don't have it working on a 2 P machine you may move that folder to a 4 P machine at some point.
 
4.  Make sure you plug the internet back in on both computers.
 
It is a bit of a pain, but I used to do it all the time to maximise my points, 6903's and 4's to my 4P systems and the 6901's to my SR2, which was better for both machines.  
 

Nice idea mmillion, but right now it may no longer matter.
Stanford's server page shows the assignment server for the 6903 and 6904 wu's (130.237.232.237) as having no more jobs. http://fah-web.stanford.edu/pybeta/serverstat.html
That's what the dark blue in the NMJ colunm to the right means.

looks full to me
1026 avail ..1026 to go
 
 

yeah, that's what it says, but those numbers haven't changed much lately: http://fah-web.stanford.edu/pybeta/logs/130.237.232.237.log.html
The important thing to me is the blue flag with a "1" in the box, which (from what i've read about and if you hover your cursor over the NMJ at the top of the column) means No More Jobs.
I still haven't got one in a while. All 8101's. My last 6903 on my 970 dropped this am (took a few days to do). Won't know what it picked up until I get home.

Yes I see that.
I just got one this morn ...a 6904
 
they also could be taking the server offline for repairs after this too
(hence NMJ's)
 
this server has had issues since January as bruce has pointed out recently
http://foldingforum.org/viewtopic.php?f=18&t=21829

Xavier Zepherious


Yes I see that.
I just got one this morn ...a 6904

they also could be taking the server offline for repairs after this too
(hence NMJ's)

this server has had issues since January as bruce has pointed out recently
http://foldingforum.org/viewtopic.php?f=18&t=21829

Lucky!! You must have been the one to drop the count from 1026 to 1025!
Yes, I had problems with that server just like the others...that's why I'm still hoping that this is only a temporary thing..

Well I did get a 6904 on my hex core 970..hmmm
Decided to OC the 4p! Working good. The 8101 frametimes should drop some.. we'll see.

I picked up a 6903 on one rig but still got an 8108 on another.

Bios flash & overclock successful! (Thanks [H])
Running a 250 refclock for 2.5GHz 
8101wu TPF dropped from 22:32 to 17.52
PPD went from 161,511 to 230,009!!
core temps in mid 30's
zero ht-retries
 
There's a little room to go higher. I'll give it a few days to make sure the OC is stable across a variety of wu's before I push the limits.
Hope I get a 6903 or 4 on this puppy to see what it can really do!
 
 

mflanaga


Bios flash & overclock successful! (Thanks [H])
Running a 250 refclock for 2.5GHz 
8101wu TPF dropped from 22:32 to 17.52
PPD went from 161,511 to 230,009!!
core temps in mid 30's
zero ht-retries

There's a little room to go higher. I'll give it a few days to make sure the OC is stable across a variety of wu's before I push the limits.
Hope I get a 6903 or 4 on this puppy to see what it can really do!
 

Nice! I notice your jumping up the Top 20 Producers.. Well Done!

from foldingforum 
  
  
by bruce » Wed Jul 11, 2012 10:35 am 
Until the 690x projects end. No matter what you prefer, the Pande Group plans to benchmark future projects more like the 8101. Do not base future hardware plans on projects 690x. Deadlines have been tightened to be consistent with a 16 physical core requirement. 

  
  
by Grandpa_01 » Wed Jul 11, 2012 7:20 am 
Sorry but that is the way PG has chosen to go for now. You need less than 16 cores to get the 69xx WU's my Gulftowns are still getting them, you could take those 2 5645's out of the sr2 and put them in single processor boards OC them and make double the PPD running 69xx WU's pretty sad situation isn't it, a single processor rig can make more PPD than a MP.


sorry for any bad news 

Well, well, how totally unshocking. PG screwing us again. But tisk, tisk... now it is affecting everyone.

it has helped in our fight with the [H]orde 
(it's hurt them worse from the looks of it)
 
yes I know it suxs, but such is life.
Im  just thankful I'll catch some before they run out

Xavier Zepherious


(it's hurt them worse from the looks of it)

You know it, considering they are almost exclusively AMD 4P systems. You know it's bad when you even get some kind of comment from Grandpa.