• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Why P6701?

mplee said:
I pretty much have them on every machine I have running. It looks like they're just trying to get as many out as possible. As of the last server stats update, there was 34.5k of them assigned.
Click to expand...
wow that's a lot of 6701s
 
Suckers....

I did not get a 6701.... I got a 6702.... I feel so special. :mad:
 
jebo_4jc said:
wow that's a lot of 6701s
Click to expand...


6702's as well.. its a cycle for me, i get 2 6701's back to back then a 6702 then another 2 6701's then a 6702.. endless cycle.. ive only had 1 non 670x WU in the last 10 days.
 
jebo_4jc said:
wow that's a lot of 6701s
Click to expand...

I'm not sure they've ever had that many out at one time. I think the netload issues the server had, prevented them from allowing it to assign that many. They have a script running on it now that's supposed to fix the issue.

I did notice yesterday that some uploads took forever. My usual speed is around ~325KB/s....it was around 1/10 of that. I'll take the super slow upload speed versus not being able to upload at all.
 
I'm bruning threw them on the SR-2 rigs.

/wonders when they will finaly run out.
 
It seems 670x will never run out.... It takes so long to complete 1 WU so Stanford would have more time to create more and more 670x WUs ... lolz
 
I just started folding again, and the first WU I pick up is a 6701. Sitting at 0% for over 20 minutes. :(
 
I'm about ready to sacrifice a chicken. (like they did in Major League though...not a live one, a KFC one) :D
"You tryin' to tell me Jesus Christ can't hit a curve ball??!" :D
 
On your 1055T ?

If so that is a sign of a problem.....it should have a 10 min TPF or so at stock, a 4.0 GHz thuban will do ~8:30
 
1055t no OC.
I'll start/stop the darn thing. brb.

-updated-
Stopped client. Deleted WU. Started client. Picked up the 6701 again. Will report back in a few minutes.
 
Shambler said:
1055t no OC.
I'll start/stop the darn thing. brb.

-updated-
Stopped client. Deleted WU. Started client. Picked up the 6701 again. Will report back in a few minutes.
Click to expand...

Deleting workunits is bad for science, please don't do it...
 
I'm getting more tired of the whining than the 670x's.
 
Xilikon said:
Deleting workunits is bad for science, please don't do it...
Click to expand...
No problem. First and last time I delete a WU.

Update: It has been about 10 minutes and still sitting at 0%. I'll let it sit for a bit longer then maybe start a thread.
 
Tobit said:
I'm getting more tired of the whining than the 670x's.
Click to expand...

I know, I know.

Do you happen to know what this project is for anyway? I can't find it in the project descriptions at the folding forum.
 
tjmagneto said:
Do you happen to know what this project is for anyway? I can't find it in the project descriptions at the folding forum.
Click to expand...
It is one of their old A2 WUs that was converted to the A3 core and is a collaboration with their Swedish colleagues to study ion channels. This was new science for PG on the A3 core as of June of this year and no other projects other than 6702 are doing any kind of ion channel studies at the moment so these are quite important from a scientific viewpoint.

I believe someone in the other 670x thread stated that they suspected the lower points is because they were benchmarked on older hardware and not the current Core i5 benchmark machine. This is not the case as these WUs came out well after they switched to the i5 for benchmarking.

I'm really not sure what the issue is with these WUs scoring lower. We experienced some of the same issues in beta testing with some people actually scoring more points on 6701's compared to some other WUs such as 6052 and 6023. Some have speculated the lower, and higher, scoring differences was due to cache sizes or HyperThreading but no final conclusions were made before these were released to gen-pop.
 
Xilikon said:
Deleting workunits is bad for science, please don't do it...
Click to expand...

What are you supposed to do in this case? (Actual question, not being smart ass)

Having a WU stuck on 0% forever can't be good for science, either.

On the 670x side of things, I think this is actually a good thing...

For now only 670x projects for everybody, so no bad luck relative to anybody else.

Maybe if they just keep getting rid of all 670x projects then people will not be stuck with them when other are getting cream units.
 
That's one of FAH's biggest mysteries to me.....how in the world can a static "benchmark machine" produce such wildly inconsistent results? If the benchmark machine's hardware is static over the course of benchmarking multiple WUs, the PPD should stay relatively the same (+/- some minor performance differences due to the impact of certain architectures). Instead we see wild fluctuations (exhibit A: the difference between 6701 and 60xx units).

I would love to know more about the benchmarking process, and how they can so baldly assert in the Points FAQ that performance should be relatively the same across all WUs (again, with the caveat of minor architectural differences).
 
SpeedyVV said:
What are you supposed to do in this case? (Actual question, not being smart ass)

Having a WU stuck on 0% forever can't be good for science, either.

On the 670x side of things, I think this is actually a good thing...

For now only 670x projects for everybody, so no bad luck relative to anybody else.

Maybe if they just keep getting rid of all 670x projects then people will not be stuck with them when other are getting cream units.
Click to expand...


if its stuck do all the usual trouble shooting before deleting.. last resort delete it. if its stuck then the WU's bad and its going to get deleted from the system anyways.
 
sirmonkey1985 said:
if its stuck do all the usual trouble shooting before deleting.. last resort delete it. if its stuck then the WU's bad and its going to get deleted from the system anyways.
Click to expand...

Yes, deletion of WU should be used as last resort only, only if it's confirmed you won't make the preferred deadline. Other steps to do would be :

-Try to move the FAH folder to another computer to see if it crunch this (if you have another SMP around).
-Delete the Core file and let it download the latest one.
-Check the log files for any suscipious entries.
-Check if your computer is fully stable, do benchmarks and stress testing to make sure.
 
Xilikon said:
Yes, deletion of WU should be used as last resort only, only if it's confirmed you won't make the preferred deadline. Other steps to do would be :

-Try to move the FAH folder to another computer to see if it crunch this (if you have another SMP around).
-Delete the Core file and let it download the latest one.
-Check the log files for any suscipious entries.
-Check if your computer is fully stable, do benchmarks and stress testing to make sure.
Click to expand...

This is good info. Thanks.

I thought the Core file was the WU. So it is OK to delete FahCore_15.exe or whatever and restart the WUs?
 
SpeedyVV said:
This is good info. Thanks.

I thought the Core file was the WU. So it is OK to delete FahCore_15.exe or whatever and restart the WUs?
Click to expand...

Yes, that's fine. Be warned that there are possibilities that having a new core can cause the WU to get trashed due to incompatibilites. If that happen, be reassured it's not you who did it willingly and hope the system pick it as a bad WU.
 
sirmonkey1985 said:
now with the bonus points its all about how many points you can get with a single system and when we lose 5-6k points while still making 60k+ on the same system we all go ape shit about it.
Click to expand...
That's one of the reasons why I don't like the bonus system as it stands in the present, and my opinion is to either revise the bonus structure (I know how to go about it), or drop it entirely and opt to assign generous static point values to -bigadv instead.
 
rmdashrrootsplat said:
670x was designed to get people with C2D's to fold more uniproc units.

/tinfoil :D
Click to expand...
Yes, my control WU theory. Like the P2684 is maintained to compel people with 'marginal' (note the quotes) systems from folding -bigadv to fold regular A3 instead. :D ;)

j/k
 
Xilikon said:
Try to move the FAH folder to another computer to see if it crunch this (if you have another SMP around).
Click to expand...

Does this actually work? Would you need to leave the original machine that downloaded the WU 'offline' until the WU is completed elsewhere or can you just start up a new client and continue on?

I tried to move a -smp WU from one machine to another once and it just downloaded a new WU instead of continuing the one I'd been working on. Perhaps I didn't do it right.
 
APOLLO said:
That's one of the reasons why I don't like the bonus system as it stands in the present, and my opinion is to either revise the bonus structure (I know how to go about it), or drop it entirely and opt to assign generous static point values to -bigadv instead.
Click to expand...

The bonus system strongly contributed to this problem, along with the 670x units poor bonus points returns.

People cherry picked and flushed out undesirable 670x units in the beginning and now all the honest folders are feeling the effects.
 
well now im kinda pissed off.. finally got something other then a 6701/2 WU and my effing hdd dies(yes i run the client on a separate hard drive then my boot drive). so now i have to reinstall the damn client again!
 
I wonder if they fixed things if it would be at all the same experience.

Right now there are a few times a day where I set my alarm, and see what the new unit will be... if it is a 6701, I swear, and reset my alarm for later. If it is a 2684, I say oh well, but at least I know I have a day without 6701s. If it is a 2685, I smile. A 2686 and you get a hells yeah, a fist pump, and a real rush.

Running two fast bigadvs side by side gives you a feelgood buzz all day. Both crunching 6701s for 24 hours straight makes you bang your head.

But what if they actually fixed things? Be careful of what you wish for - I can't image getting the same points every day, it would be like watching paint dry. It would be the MTV generation - we feel neither highs nor lows (and how does that feel?) Meh. :p
 
Aix. said:
Does this actually work? Would you need to leave the original machine that downloaded the WU 'offline' until the WU is completed elsewhere or can you just start up a new client and continue on?

I tried to move a -smp WU from one machine to another once and it just downloaded a new WU instead of continuing the one I'd been working on. Perhaps I didn't do it right.
Click to expand...

You must move the whole FAH folder including the subfolders to be able to resume. I think you moved only a part and since there is no work files found, they will pick a new one.
 
The i7 980 has only seen 6701, 6702, and 6052 since it's been Folding. But it's still doing nicely with them.
 
was wondering if it would be faster to run a -nofred diskless as I'm going to be away for 7days...
 
cleanncut said:
was wondering if it would be faster to run a -nofred diskless as I'm going to be away for 7days...
Click to expand...
from what I understand, Windows smp folding is faster than Linux these days.
 
may i be worthy said:
But what if they actually fixed things? Be careful of what you wish for - I can't image getting the same points every day, it would be like watching paint dry. It would be the MTV generation - we feel neither highs nor lows (and how does that feel?) Meh. :p
Click to expand...
Well, from the perspective of a person that has been folding before the nVidia GPU days, I could say it would definitely be less interesting on a day to day basis. You won't see anything like the sharp spikes we see today because of -bigadv or the bonus system in general. The big *but* here is that we will return to some semblance or 'normality' with superior representation of reward for research.

Look at it from a different angle, I have over 20 clients that are a mix of GPU, regular and -bigadv SMP (no uniprocessor clients). My output is equal to the output of some people running a single SR-2 with the fastest config and folding the most productive -bigadv WUs. Now, if someone doesn't see that there is something skewed about the bonus system and how it greatly exaggerates reward for the highest caliber systems out there, there's not much I can say to convince them otherwise.

I know all the arguments pro high bonus structure such as the low power to production ratio, low maintenance vs multiple client numbers and types, the recompense for high monetary investment, ostensible commensurate research compensation, etc., and I have a counterargument for each that encompasses both the folding perspective as well as the ethical side of the debate. In truth, I probably have benefited one of most on this team from the -bigadv bonus because I have been accruing its benefit for a full year, yet I am one of its biggest opponents as it is being implemented today. We need to see the full picture, not only short term gains.
 
Last edited:
You must log in or register to reply here.
Back
Top