@Jedi. Thx for your time and input. I applied the team number (doh had forgotten before, was just on the Client settings tab). Then everything was fine. As normal :D
 
Thank you. Awesome program. Hasnt given me any problems, neither smp, big adv or GPU2 folding, on 3 machines :D

Ok, Am i doing something wrong??? I switched over to this outstanding program, an i love it!!! But I have noticed that my CPU runs cooler (using realtemp) 60 - 65 * each core, as when I ran just the smp long install and setup version I ran 70 to 75 * each core, also noticed it takes about three to four days to finish a wu and it used to take 24 hours to 2 at the worst 3 days...
 
So what did I do wrong???

Jedi, just a note of thanks for this great setup program.  Just a little concerned that the point totals I see in the tracker aren't matching what folding@home has.  Could the smp bonus be wrong or something?

You have to complete 10 WU's with the passkey to get the bonus for SMP wu's.

 

baronmind


Jedi, just a note of thanks for this great setup program.  Just a little concerned that the point totals I see in the tracker aren't matching what folding@home has.  Could the smp bonus be wrong or something?

 
There is no way to figure out if you have completed the 10 WUs required to start getting bonus points, so the Tracker assumes that bonus points will be awarded for every SMP A3 WU returned within the preferred deadline. This can cause the Tracker to show a much higher value for total points compared to your official stats until you complete the 10 WUs.
 

barnhartmd051977


Ok, Am i doing something wrong??? I switched over to this outstanding program, an i love it!!! But I have noticed that my CPU runs cooler (using realtemp) 60 - 65 * each core, as when I ran just the smp long install and setup version I ran 70 to 75 * each core, also noticed it takes about three to four days to finish a wu and it used to take 24 hours to 2 at the worst 3 days...

So what did I do wrong???

 
Is your CPU load 100% as expected? If not then that can be a few things:
 
1. The wrong -smp x flag is being given to the client (post FAHlog.txt if this is the case)
2. The client may not be getting the proper affinity settings (check in task manager if it is set to all CPUs)
 
If the CPU is at 100% load there are a couple more things to check:
 
1. Is there a process other than FAH using a lot of CPU? If so then that is the problem.
2. If you have checked all of the above, you can try deleting the current WU to see if it's just a bad WU.

i have been folding for a while. i just switched to the tracker program, and let it re download all of my clients. do i have to do 10 A3 WU again to start getting bonus points or does it just keep track with your passkey so changing clients wont matter?

^^As long as you use the same passkey as your other clients, it'll be fine.

Version 3.31 has been released to fix a few bugs.

bad WU thanks Jedi95
 

Just installed this 3.31 FAH tracker and saw a 50% jump in productivity!
 
I am a new folder who has been at it for almost 2 weeks now with my new GTX470.  I was having a little trouble with several seperate clients, so I said to hell with it and decided to try this. IT downloaded the proper clients for my system, detected my GTX470 and I7-950 right away and within 5 minutes of setup I was away and folding with a double in PPD than I was experiencing prior thanks to being able to get my Quad core CPU folding on the right client.
 
I am now banging out 21,000 PPD. In a word, Awesome! Now If I can only figure out how to OC the GTX470 a little higher I might be able to hit 25k/Day.  Once I add that second video card (thinking about another 470 or maybe even a GTX480, thoughts anyone?) maybe I can get this machine up to 40,000+PPD.
 
Curing diseases and warming my house at the same time lol.
 
Recommended!

Version 3.32 has been released to fix several bugs.

I haven't gotten a bigadv since I set up GPU tracker ( I knowtheres a shortage) but I wanted to make sure. I just clicked the bigadv box, I didn't select -advmethods or any of the other boxes.

tank1023


I haven't gotten a bigadv since I set up GPU tracker ( I knowtheres a shortage) but I wanted to make sure. I just clicked the bigadv box, I didn't select -advmethods or any of the other boxes.

 
-advmethods is automatically removed from the SMP client if -bigadv is selected. As long as you applied the settings, the next time the client starts it should have the -bigadv flag given. The shortage is the most likely cause of this. Both my i7 systems are running bigadv WUs right now through the Tracker without problems.

Hey jedi... in my thread about 912/925 WU's there is now info that in order to get those WU's one must have -advmethods tag in place.  Looking at my setup the Tracker has the following arguments in place for my Fermi cards:
 

 Arguments: -oneunit -forceasm -forcegpu nvidia_fermi -gpu 0    
 Arguments: -oneunit -forceasm -forcegpu nvidia_fermi -gpu 1   

 
Is this something you can look into to see if we can get that argument/tag added?
 
Here's the post I mentioned above...
 
http://www.evga.com/forums/fb.ashx?m=652210
 
TIA,
 
JD

JD22964


Hey jedi... in my thread about 912/925 WU's there is now info that in order to get those WU's one must have -advmethods tag in place.  Looking at my setup the Tracker has the following arguments in place for my Fermi cards:

 Arguments: -oneunit -forceasm -forcegpu nvidia_fermi -gpu 0     
  Arguments: -oneunit -forceasm -forcegpu nvidia_fermi -gpu 1   

 
Is this something you can look into to see if we can get that argument/tag added?

Here's the post I mentioned above...

http://www.evga.com/forums/fb.ashx?m=652210

TIA,

JD

 
This is already an option, just check the box here:

jedi95

This is already an option, just check the box here:

Doh!
 
Sorry about that...
 

Hi jedi95,
One question regarding Games.txt: Will this file be updated frequently? I've added several games that don't appear in the default listing and want to know if I should expect to append those games each time there's a new release.
 
And an observation: I have a 295 with one core running hotter than the other so I have different heat control entries for each GPU core. I've noticed on the last two bug fixes (as long as I've been running your application) only one of the two entries migrate into the new version. I'll keep an eye on this and forward specifics after the next release.
 
The way I use it: I set a modest default GPU usage percentage per core then monitor temperatures by WU points increasing GPU usage percent up to what I consider a safe temperature limit.
 
Last but not least: I'd like to add my thanks for your app and the time/effort you've put into it.
 
Chris

Evil_Monkey_153


Hi jedi95,
One question regarding Games.txt: Will this file be updated frequently? I've added several games that don't appear in the default listing and want to know if I should expect to append those games each time there's a new release.

And an observation: I have a 295 with one core running hotter than the other so I have different heat control entries for each GPU core. I've noticed on the last two bug fixes (as long as I've been running your application) only one of the two entries migrate into the new version. I'll keep an eye on this and forward specifics after the next release.

The way I use it: I set a modest default GPU usage percentage per core then monitor temperatures by WU points increasing GPU usage percent up to what I consider a safe temperature limit.

Last but not least: I'd like to add my thanks for your app and the time/effort you've put into it.

Chris

 
My method of updating Games.txt no longer works since the newer Nvida drivers use a different method of storing the list of games. I have to find some other way of doing this now.
 
For Heat Control, make sure it saves the settings and that you are using the GPU2 client. The settings for Heat Control only save when you switch which GPU you are changing settings for or hit apply. Heat Control also doesn't work with GPU3 clients (core 15 ignores the environment variable)

jedi95

For Heat Control, make sure it saves the settings and that you are using the GPU2 client. The settings for Heat Control only save when you switch which GPU you are changing settings for or hit apply. 

 
Understood. I know they're being saved because the client has to be stopped and restarted to apply the heat control settings - I'm monitoring these in real time.  
 
I'll keep the GPU3 client limitation in mind when I make the leap.
 
Chris

hey jedi. still confused. I'm running -bigadv and want to keep it selected. if I select -advmethods (for my gts 450's) what will happen to the -bigadv on my i7?
 
Also, if I want to switch the number of cores between 7(while at home) and 8 (while im at work/sleep) do i have to select it and then stop all my folding, exit and restart? What do i need to do to switch between 7 and 8 cores?

Carlitos714


hey jedi. still confused. I'm running -bigadv and want to keep it selected. if I select -advmethods (for my gts 450's) what will happen to the -bigadv on my i7?

Also, if I want to switch the number of cores between 7(while at home) and 8 (while im at work/sleep) do i have to select it and then stop all my folding, exit and restart? What do i need to do to switch between 7 and 8 cores?

 
If you select -advmethods without also selecting -bigadv then the SMP client will have -bigadv. If you select both -advmethods and -bigadv the SMP client will still use just -bigadv. Basically the -advmethods flag won't do anything to the SMP client if -bigadv is enabled.
 
The SMP cores setting takes effect the next time the SMP client starts. This will happen automatically the next time it starts a WU, or you can apply it immediately by stopping/starting the SMP client.

jedi95

Carlitos714


hey jedi. still confused. I'm running -bigadv and want to keep it selected. if I select -advmethods (for my gts 450's) what will happen to the -bigadv on my i7?

Also, if I want to switch the number of cores between 7(while at home) and 8 (while im at work/sleep) do i have to select it and then stop all my folding, exit and restart? What do i need to do to switch between 7 and 8 cores?

If you select -advmethods without also selecting -bigadv then the SMP client will have -bigadv. If you select both -advmethods and -bigadv the SMP client will still use just -bigadv. Basically the -advmethods flag won't do anything to the SMP client if -bigadv is enabled.

The SMP cores setting takes effect the next time the SMP client starts. This will happen automatically the next time it starts a WU, or you can apply it immediately by stopping/starting the SMP client.

 
Thank you very much for the explanation!!!!!

Hi,
 
My bigadv bonus has gone, now this is the second time that i stop all the clients to restart the computer, and when GPU Tracker starts all again the bonus its gone, what could be causing this?

A quick note regarding my heat control issue pointed out earlier: the HeatControlSettings.txt file within the GPU0 folder has been deleted twice; once overnight and a second time this morning. I've noticed this while monitoring GPU temperatures through EVGA Precision. The corresponding file in GPU1 has remain unchanged and I've copied it back to GPU0 to reinstitute my heat regulation rules.
 
I'm saving all my log files so they're available to you should you wish to review them.
 
Chris

Dont know if this is a bug or something in the latest version, but when running the SMP client in oneunit mode it will complete a WU but then wont turn off completely - it will not start another unit but still needs to be manually stopped as the PPD box stays yellow at 0.  When you stop the CPU it says error stopping CPU code=5. 
 
This might just be me, but I am pretty sure it started after one of the recent updates, I cant remember which one as I dont run SMP often.  Wonderful program tho, thanks Jedi!

Al_Ho


Dont know if this is a bug or something in the latest version, but when running the SMP client in oneunit mode it will complete a WU but then wont turn off completely - it will not start another unit but still needs to be manually stopped as the PPD box stays yellow at 0.  When you stop the CPU it says error stopping CPU code=5. 

This might just be me, but I am pretty sure it started after one of the recent updates, I cant remember which one as I dont run SMP often.  Wonderful program tho, thanks Jedi!

 
This has already been fixed and it will be out in the next version.

I just used this for the first time this weekend and while it still did not make my 450 fold nice with other cards(not the trackers fault) it was the easiest and fastest set up ever.
 
great job on a great app
 
thanks jedi

Question for you Jedi.
 
Today, I already had a SMP CPU job (a 6701 WU) running on my PC that was ~ 32% complete via the "FAH-630.exe" program.  I wanted to play with your FAHGPU app from the perspective of running both that CPU job and a couple of GPU WU's on my GTX470's. 
 
So I first temporarily suspended the 6701 CPU job via Cntrl C within the FAH-630 client and it closed normally.  Then I started up your FAHGPU software and enabled the SMP and both GPU clients to run.  Within the first 5 mins, the GPU clients began running and reporting their % Completion as expected.  The CPU client was also running and reported that it was working on a 6701 WU at 0% Complete.  I watched for the progress of the CPU client to report the first amount of progress and when it did report, it showed 1% Complete.  I was expecting that it would grab onto that uncompleted 6701 WU that I had previously suspended (which was at ~ 32% Completion when I suspended it).  But, FAHGPU showed 1% Complete instead. 
 
So, my question is, should FAHGPU have picked up on that previous 6701 WU and reported ~ 32% Complete rather than 1%, or did it start a new 6701 WU?
 
I'm also running the latest version of your software.
 
Thanks,
 
Rick

texinga


Question for you Jedi.

Today, I already had a SMP CPU job (a 6701 WU) running on my PC that was ~ 32% complete via the "FAH-630.exe" program.  I wanted to play with your FAHGPU app from the perspective of running both that CPU job and a couple of GPU WU's on my GTX470's. 

So I first temporarily suspended the 6701 CPU job via Cntrl C within the FAH-630 client and it closed normally.  Then I started up your FAHGPU software and enabled the SMP and both GPU clients to run.  Within the first 5 mins, the GPU clients began running and reporting their % Completion as expected.  The CPU client was also running and reported that it was working on a 6701 WU at 0% Complete.  I watched for the progress of the CPU client to report the first amount of progress and when it did report, it showed 1% Complete.  I was expecting that it would grab onto that uncompleted 6701 WU that I had previously suspended (which was at ~ 32% Completion when I suspended it).  But, FAHGPU showed 1% Complete instead. 

So, my question is, should FAHGPU have picked up on that previous 6701 WU and reported ~ 32% Complete rather than 1%, or did it start a new 6701 WU?

I'm also running the latest version of your software.

Thanks,

Rick

 
The Tracker can't import your existing clients or WUs, so everything starts from 0%. The Tracker is independent of any existing FAH clients you have installed, so it won't import or modify them in any way.

I'm no expert on how Stanford handles jobs started on different client software.  So, just for clarification, was the 6701 WU that FAHGPU initially started a totally new 6701 job or was it the existing 6701 job that I had started and was 32% complete as far as Stanford is concerned.  I'm just needing to know if I have another 6701 (that FAHGPU started) hanging out there to complete or was the one that FAHGPU reported the same one that I had running (just listed at 0% complete).  Hope this makes sense...
 
Thanks for the help,
 
Rick