P6901 - new Swedish bigadv

whaaa.. a work unit with a description.. can't be it must be a mistake.. :p lol

only 3 years of bitching and they finally made a description for a work unit thats actually understandable..
 
First frame done on each machine - frames times absolutely identical to p6900, so no performance improvements yet.
 
sirmonkey1985 said:
hey at least they aren't lower. so that's still a good sign.
Click to expand...

Yeah, good point. I sometimes worry that our endless campaign to get them to fix the 2684 points will be "fixed" by reducing every other project to match, rather than raising 2684 to them.

It looks like we won't have bigadv drought problems for the next few months at least.
 
Here is what worries me. The deadline on the 6901 according to that description is 2 days. If the frame times are still the same as a 6900, that means my Bloomfield is no longer able to do bigadv and my Gulftown will just barely make it.
 
Hmmm, HFM is reporting a deadline of 4 days, and the description says 3 days, with preferred of 2 days. Something is up.

HFM shows Client type = unknown, but it does record Core as GRO-A5 and Core ID as A5, and gets PPD correct.
 
amdgamer said:
Here is what worries me. The deadline on the 6901 according to that description is 2 days. If the frame times are still the same as a 6900, that means my Bloomfield is no longer able to do bigadv and my Gulftown will just barely make it.
Click to expand...

they futzed ... its still 4 and 6

at least according to the servers you turn them into...
http://fah-web.stanford.edu/psummary.html
 
Patriot said:
they futzed ... its still 4 and 6

at least according to the servers you turn them into...
http://fah-web.stanford.edu/psummary.html
Click to expand...

I hope so, but maybe that was a message of things to come from Kasson. Maybe his subconcience outed his future plans accidently via that futzed product description for the new 6901's.

Either way, i'm going to stay behind and do A3's for now until bigadv dries up. Someone has to mop up all those 2684's that are probably still laying around :D
 
amdgamer said:
I hope so, but maybe that was a message of things to come from Kasson. Maybe his subconcience outed his future plans accidently via that futzed product description for the new 6901's.

Either way, i'm going to stay behind and do A3's for now until bigadv dries up. Someone has to mop up all those 2684's that are probably still laying around :D
Click to expand...

lol... those evil units... please take them away... they give me nightmares ....
 
Patriot said:
lol... those evil units... please take them away... they give me nightmares ....
Click to expand...

I actually love those 2684's because they make my bigadv upload schedule more predictable. I hate it when you are in the middle of a game when folding 1 or 2 decides it wants to upload a finished work unit. It kind of shuts down things, which is why I upgraded to the 30Mbpsdown/5Mbpsup extreme service from TimeWarner.
 
amdgamer said:
I hope so, but maybe that was a message of things to come from Kasson. Maybe his subconcience outed his future plans accidently via that futzed product description for the new 6901's.
Click to expand...
I think it's quite unlikely that they will reduce the deadlines for new -bigadv units. The current ones seem quite well-adjusted for the most part; if they made them shorter then that would significantly reduce the number of machines capable of finishing units in time.
 
I7 950 @ 3.6 - 34:37/frame, 30K ppd. Sounds about the same as other bigadv, although I have no history on this machine...its very first bigadv... :)

Edit: It is staying around 34:40 after 5 frames.
 
Last edited:
Zero82z said:
I think it's quite unlikely that they will reduce the deadlines for new -bigadv units. The current ones seem quite well-adjusted for the most part; if they made them shorter then that would significantly reduce the number of machines capable of finishing units in time.
Click to expand...

You are right, but I often wonder what is going to happen with the regular non bigadv stuff as more and more systems become bigadv capable. I wonder whether the PG will end up with a massive backlog like they have with the uniprocessor classic work units.
 
If it gets to that point, they'll probably just get rid of the -bigadv flag entirely and mix SMP and -bigadv projects together.
 
musky said:
I7 950 @ 3.6 - 34:47/frame, 30K ppd. Sounds about the same as other bigadv, although I have no history on this machine...its very first bigadv... :)
Click to expand...
sniff they grow up so fast
 
Also right on the money with P6900 frames times on a dual X5550 machine @ stock and a dual hex SR-2 @ 3.6.
 
I've not seen any yet. Still getting a spattering of P2684, P2685, & P2686's.
 
Currently I'm using -smp 11 with bigadv and GPU folding so it looks like I need to try other values once I drop in the binary. From what I can gather from this posting by PantherX over at Folding Forum it looks like -smp 11 and other prime numbers won't be allowed. The comment was in response to another posting where the client folding client rolled itself back the from -smp 11 to -smp 10.

Actually, it is to ensure that if the donor enters a "bad" value (11, 23, etc) the client will not give errors, instead it will be "intelligent" to choose the next lower "good" value which will not give errors and can fold the WU without further issues. You can search he Forum about this if you want further details.
Click to expand...
 
tjmagneto said:
Currently I'm using -smp 11 with bigadv and GPU folding so it looks like I need to try other values once I drop in the binary. From what I can gather from this posting by PantherX over at Folding Forum it looks like -smp 11 and other prime numbers won't be allowed. The comment was in response to another posting where the client folding client rolled itself back the from -smp 11 to -smp 10.
Click to expand...

now they need a forcesmp option lol... for those that find this new intelligence lacking...
 
Agreed. SMP 11 works fine for me from a stability standpoint.

However, I recently did drop down to SMP 10 on my gulftown because my SMP and dual GPU clients were having a CPU cycle fight.


 
jebo_4jc said:
Agreed. SMP 11 works fine for me from a stability standpoint.

However, I recently did drop down to SMP 10 on my gulftown because my SMP and dual GPU clients were having a CPU cycle fight.
Click to expand...

How has that worked out for you PPD wise. I'm only folding with one of my 2 GPUs right now but I'd like to fire #2 up again. It's not folding right now as the CPU is taking a lot of hot air in from the card closest to it so it's off until I get the W/C installed.

@ Patriot: "intelligence." LOL
 
The X5650 gulftown at ~3.6ghz is getting 42k ppd on a 6900 running smp -10

The GTX470s are both at full capacity. They're producing 12k ppd at stock clocks and 14k overclocked.
 
may i be worthy said:
Yeah, good point. I sometimes worry that our endless campaign to get them to fix the 2684 points will be "fixed" by reducing every other project to match, rather than raising 2684 to them.

It looks like we won't have bigadv drought problems for the next few months at least.
Click to expand...


they won't touch the 2684 points. they will just leave it til the project ends then rewrite the WU in need be. but they won't change a WU's points in mid run since its not fair for everyone else. same thing happened with the 6701/02's. they thought about changing them a while back. but they had already been running for 6 months at that point and decided not to.
 
You must log in or register to reply here.
Back
Top