jkefalas

7 i m

stonerhino

I have no intention of folding on my 660ti until PG gets their heads out of their posterior and actually DO something.

How ironic that you suggest PG is doing nothing in the same thread that mentions 3 fahcore version updates.    (v2.22 original, v2.24 new, v2.25 newest)

I think it was implied that they need to do something more sensible than just throwing unstable WU's at us.

Forever have they been saying that larger WUs are coming.  Current overclocks may not be stable on larger and more complex work units.  Doesn't make the WUs unstable.  Makes the current overclock unsustainable.  If you have been folding a long time, then you know the same thing happened with CPU clients a long time ago.
 
Pande Group has to wait for a handful of things to line up before they can add Kepler optimizations.  www.gromacs.org actually writes the code used in the fahcores.  NV has to have updated and working drivers.  NV didn't release CUDA 5 (with full Kepler optimizations) until just this week.  And then, PG can try to work all that spaghetti in to a workable Kepler core, all without breaking Fermi or G80 GPU support.
 
Unfortunately, the most sensible thing is to do is wait patiently, which is the hardest thing to do.  It also looks like a lack of progress from PG, or even as some here would claim, GPU neglect, but it absolutely is not neglect.  No news is just that. No news.  It doesn't mean that things aren't moving forward behind the scenes.  PG just doesn't share their day to day activities, nor would they have the time to do that.
 
Ya, sure, PG should throw GPU folders a bone once in a while in the Blog or forum.  But I don't get offended if they don't do it for a while.  PG actually has a lot more interest and investment in speeding up the folding, like adding Kepler optimizations, than the cost of a few kepler GPUs.

-ZS-Carpenter

7 i m

jkefalas

7 i m

stonerhino

I have no intention of folding on my 660ti until PG gets their heads out of their posterior and actually DO something.

How ironic that you suggest PG is doing nothing in the same thread that mentions 3 fahcore version updates.    (v2.22 original, v2.24 new, v2.25 newest)

I think it was implied that they need to do something more sensible than just throwing unstable WU's at us.

Forever have they been saying that larger WUs are coming.  Current overclocks may not be stable on larger and more complex work units.  Doesn't make the WUs unstable.  Makes the current overclock unsustainable.

You must be either joking or delusional because there have been loads of unstable GPU WUs that wont run at stock clocks or underclocked for years. Overclocking is always a gamble but for WUs to not even run at stock speeds just shows how far on the back burner GPU work is to PG. It is so blatantly obvious anyone who GPU folds to any extent has seen it and its been going on for years.

ZS I'd have to agree with you. Just this week I got 4 GPU work units back to back that failed with in 0 to 10 frames.
First thought was my video card was failing or another problem. On the 5th WU "different form the other 4" its been folding and hasn't lost not 1 more WU since those 4. So yes a lot of bad GPU work units get passed out & for many years its been that way.

I try to read just about everything that Dr. Pande writes about FAH events and progress.  Can't say that I always understand some of the technical aspects, but I try to stay abreast of what we are all doing in our homes every day.  I'm reminded of the oft quoted thing in Sales.  You need an "elevator speech" that can easily and quickly communicate to someone the reasons they should consider your product.  That same concept applies to Folding when we talk to others and try to convince them to Fold.  If we can't easily and succinctly point them to what Folding is actually doing (or more importantly has done), then they glaze over and likely will not consider doing it. 
 
I do have a problem referring to that "thing" that Folding has accomplished that an average person could grasp.  I could try to "parrot" the things that Dr. Pande writes, but a lot of that stuff is about simulations and scientific geek-speak.  What quickly convinces people is being able to tell them something tangible that Folding has actually done.  Something that they can relate with.  I may have missed it, but I don't have that "elevator speech" worked out very well and wish I did.  I can't point to a disease being solved or major break-through that Folding actually provided.  Sometimes, I feel that I'm Folding completely out of "faith" that our efforts will actually one-day come to some tangible fruit.  I hope that my speaking frankly won't cause any concern for existing Folders, but Jkefalas' post caused me to stop and think about this for a moment.

texinga

I try to read just about everything that Dr. Pande writes about FAH events and progress.  Can't say that I always understand some of the technical aspects, but I try to stay abreast of what we are all doing in our homes every day.  I'm reminded of the oft quoted thing in Sales.  You need an "elevator speech" that can easily and quickly communicate to someone the reasons they should consider your product.  That same concept applies to Folding when we talk to others and try to convince them to Fold.  If we can't easily and succinctly point them to what Folding is actually doing (or more importantly has done), then they glaze over and likely will not consider doing it. 

I do have a problem referring to that "thing" that Folding has accomplished that an average person could grasp.  I could try to "parrot" the things that Dr. Pande writes, but a lot of that stuff is about simulations and scientific geek-speak.  What quickly convinces people is being able to tell them something tangible that Folding has actually done.  Something that they can relate with.  I may have missed it, but I don't have that "elevator speech" worked out very well and wish I did.  I can't point to a disease being solved or major break-through that Folding actually provided.  Sometimes, I feel that I'm Folding completely out of "faith" that our efforts will actually one-day come to some tangible fruit.  I hope that my speaking frankly won't cause any concern for existing Folders, but Jkefalas' post caused me to stop and think about this for a moment.

 
Well put, Tex.
 
I know PG does good work because I like you have taken the time to seek out what information they do release. That being said, they could stand to learn a lot from some other projects as to how to stay in touch with their donor base. I mean WCG sends me e-mails to tell me about new projects, webcasts, and milestones....that's what PG needs to do. Let people know what they are doing is paying off in the labs.
 
Before anyone says they're too busy doing science to run a PR campaign they need to stop and think where they are getting all the results that fuel the research. The answer? Their donors. Event he smallest effort in keeping the people at the other end of their wu's in the loop as to what is going on and how their donations/efforts are being used to make advancements in research could yield huge returns. 

rklapp

If Folding and FoldIt aren't your cup of tea, there's always EteRNA.

http://www.wired.com/wiredscience/2012/07/ff_rnagame

Seems fun and educational!  Unless I'm missing something though, how does it equate to distributed computing research?

Xavier Zepherious

OK!... the location of the cores in v7

http://forums.evga.com/fb.ashx?m=1764899

after you finish a WU and before you start another (so finish out your WU's)

in win 7 -> use the start button

start button->All programs-> FAHclient->DATA directory 

this will bring up windows explorer (in the FAHClient data directory)

you then go to 
cores\www.stanford.edu\~pande\win32\AMD64\NVIDIA\Fermi\core15.fah 

just delete the core file within (or rename it- ie the extension like core15.exe to core15.ex1)
   
it should download the new one. provided you have the gpu slot configured for Beta , once you start folding again

to config the gpu slot you select config
highlight gpu and press edit
select add (in extra slot options) and type in "client-type" and "beta"

besides I think the beta config starts a beta directory for it now with v7.2.9 If you use the new beta client
cores\www.stanford.edu\~pande\win32\AMD64\NVIDIA\Fermi\beta\core15.fah
---------------------------------------------------------

Yea I suffer from the new A4 WU's too...30-34k

So all I had to do was rename the file? What did that actually do? Whatever it did... I'm back in business, GPU at 75% power and running 21k PPD. will I have to manually replace that file every time?

cmoscarell


So all I had to do was rename the file? What did that actually do? Whatever it did... I'm back in business, GPU at 75% power and running 21k PPD. will I have to manually replace that file every time?

 
What did that do? - the v7 client can't see the core file(lost corrupt) and will download the new core
 
v7 will download cores when they need them. you already had the core15 (2.22) so it wasn't gonna get the new one unless a WU specified a specific core version.
so by removing/renaming the old one you forced it to fetch the latest version
 
no you do not manually have to replace the file for each WU
you will if a new core version comes out and we're in the same boat
 

In that thread he responded to you...
 
The new version of the core should be available for advanced or released fah projects but it won't overwrite the old version yet. For those who might not see this suggestion, the automatic update should begin soon.

Don't ask to join the beta team unless you really do want to monitor your clients closely and give detailed reports of any anomalies.
 
Not to clear for me. Seems they are very much knee deep in the BETA testing still for Kepler. Is that a correct interpretation of his comment? Or is it BETA in general for any new client and Kepler is about to go native/live?
 
I am so excited 

Better backup your old cores for older cards. With my testing the newer core is much slower on older video cards by just over a min. a frame. example: 2:00 old core 3:15 new core TPF.

patchesanook

will jedis,s tracker update to this?
i am computer dumb sometimes.

it might. i would just to be safe, click the update cores option if you notice a failed wu.