What should I be getting?

[W.Feather]

Ive got a AMD Phenom II X6 1090T(clocked at 3.2GHZ till it cools down), a PNY 260 c216(clocked at 576/1242/999), and am getting 12k ppd for both......is this about right?

I had it run for 3 weeks straight, and got 12k ppd on average for those three weeks, i dont know if that is about where I should be, I woulda thought that my x6 would get more than 10k ppd(I assume a 260 gets more than 2k..), just wondering, do i need to let it run for a good deal longer to get a decent idea on the PPD I get?

 

[404 - File not found]

The 260 should get at least 7kppd
the X6@3.2 about 8kppd with bonus?

12kppd sound a bit low, could be the luck of the draw with the work units (some units give more points)
If you are using Fahmon to calculate your ppd it doesn't add bonus points, get HFM.

 

[W.Feather]

ive been looking on hardfolding.com

figured after three weeks it would level off the numbers for the averages

highest I saw it go(checked while away) was 12.5k.....not doing bigadv(6 cores isnt enough, IIRC i read that 4.0 would be whats needed on an AMD hex to do bigadv and make deadlines)......what is HFM? still a folding n00b

 

[Zero82z]

Get HFM.NET here and use it to monitor your clients. What PPD values are you getting for each of the clients individually? Also, have you folded ten or more SMP work units with a passkey so you qualify for the bonus points system?

 

[W.Feather]

I shuold have done atleast 10 units on the CPU, over 300 for both according to hard folding....ill put on HFM here in a min and edit with what it says i'm getting......dont know how to tell on hard folding

 

[W.Feather]

well I got it, downloaded, installed.....messing around with it now....gotta figure out how to link it to my F@H, thinking its just directing to the F@H config file, going to give that a go now

edit: just found what I think is the F@H Config.....can only import a .txt, do i need to make a copy of it in a .txt format?(i'm tryin now)

edit2: well that didnt work......can someone give me a quick down and dirty explanation(can skip simple stuff) on what to do to hook this up? Do i need a new Client to be set up?

 

[amdgamer]

The other thing you can do is go over to http://www.linuxforge.net and use his SMP2 calculator. All you need to do is plug in your current TPF's for the current work unit, and it will give you an estimate of the ppd averages based on that work unit. I myself don't even bother with any of these monitoring programs, but i'm also only running 2 rigs vesus some of the guys who are running a farm.

 

[W.Feather]

ill look at that tomorrow(doing photoshop work.....stiching together tons of images seems to not like my PC very much ), and anyhow i intend on running a mini farm (nothing on the scale, but every bit helps), so mass monitoring program wont be too bad to have

 

[404 - File not found]

well I got it, downloaded, installed.....messing around with it now....gotta figure out how to link it to my F@H, thinking its just directing to the F@H config file, going to give that a go now

edit: just found what I think is the F@H Config.....can only import a .txt, do i need to make a copy of it in a .txt format?(i'm tryin now)

edit2: well that didnt work......can someone give me a quick down and dirty explanation(can skip simple stuff) on what to do to hook this up? Do i need a new Client to be set up?

you need to create a new config then Client - Add Client. Copy paste the path of your Folding folder, similar to this:
CPU:
C:\Program Files (x86)\Folding@Home Windows SMP Client V1.01\
GPU:
C:\Users\Remi\AppData\Roaming\Folding@home-gpu\

 

[Zero82z]

edit2: well that didnt work......can someone give me a quick down and dirty explanation(can skip simple stuff) on what to do to hook this up? Do i need a new Client to be set up?

In HFM, go to Clients>Add Client. Give it a name, click the button next to the log folder box, and select the folder your client is installed in. If it's the SMP client and you followed the old version of my guide, the correct directory would be C:\Program Files (x86)\Folding@home Windows SMP Client V1.01. Once you've selected the directory, click Ok and it should be good. For the GPU client, if you're using the console version then the directory will be wherever you put it. If you're using the system tray client, then it will be located in C:\Users\*your user name*\AppData\Roaming\folding@home-gpu.

 

[W.Feather]

k, i just followed what you said exactly, and , and cant get it to run....do I need to change what the MHz is?

when I double click on the client int he window it takes me to the folder that I put the log folder for where the client is....any ideas?

 

[Zero82z]

What do you mean by not being able to get it to run? You don't need to modify the MHz field. Double-clicking on a client is supposed to take you to the folder the log file is in. That's not important. What's important is the information HFM gives you about your running clients.

 

[W.Feather]

well i guess what i mean is that i cant start it in HFM.NET, when I do that it opens the other window, understandable and I figured it would so it can read the info there, but HFM does not seem to recognize it is there

 

[W.Feather]

interesting, this time i let it wait to complete another percent before I refreashed the client table, and it saw it, last time i didnt......

now when i try the GPU client, it is broken, missing a .dll for some reason, cudart.dll to be exact


and as of now, its estimating 3658.8 ppd for my X6.....ill turn it on again before I go to sleep, and see what it says in the morning......its a P6051 wu

 

[404 - File not found]

just to make sure you understand, HFM only reads the log files from your different clients, then estimates the points you'll receive if you keep your current speed.
You must first start your clients normally. HFM only tells you the points you're getting and if any client is having difficulties.

Doublecheck the path of your GPU client, if problem persists, try reinstalling Folding gpu client. If it's still not working, reinstall nvidia drivers.

3658 ppd is low if the client has been running nonstop.

 

[W.Feather]

heres the recent client(smp) work run:

3 weeks non stop(21 day, 20 hour)
stopped last night at 9ish EST, forgot to turn on again after game till an hour from now.


i would assume it watches the last three weeks logs....

 

[sirmonkey1985]

interesting, this time i let it wait to complete another percent before I refreashed the client table, and it saw it, last time i didnt......

now when i try the GPU client, it is broken, missing a .dll for some reason, cudart.dll to be exact


and as of now, its estimating 3658.8 ppd for my X6.....ill turn it on again before I go to sleep, and see what it says in the morning......its a P6051 wu

go to your users/username/appdata/roaming/folding@home-gpu.. take the DLL files and copy those then put them in the main folding gpu folder in your root directory..

your ppd will change because if you stopping the client now that its going to complete later then before your bonus points will be lower for that WU thus your PPD drops.. next WU it should go back to normal.. make sure you have HFM showing PPD for last 3 frames as well..

 

[404 - File not found]

i would assume it watches the last three weeks logs....

HFM uses the last three % actually (and Hardfolding.com calculates over a week's time)
You can change this under Preferences - Options - Calculate PPD based on: AllFrames
It should show you more points now.

Sadly the way the smp works, pausing a few hours hurts the bonus points you receive.

 

[Zero82z]

i would assume it watches the last three weeks logs....

HFM only reads the frame times since the last time the client was started. As an example, if you run a unit to 50%, exit the client, and then start it back up again, HFM will not look at the times from the first 50% and will instead count only the frames you've completed since you loaded up the client again.

 

[W.Feather]

ahh, that explains it.

got the GPU working.......I didnt realize that my GPU client was starting with windows.....so it tried opening multiple copies of the GPU client, which caused the error there.

edit, i just looked at core temp and saw something interesting.....no core has more than a 70% load going on it......could HFM have something to do with this? Yesterday when I ran this, before the new unit I was sent, when I watched it was doing the normal load the cores to 100%...

 

[sirmonkey1985]

ahh, that explains it.

got the GPU working.......I didnt realize that my GPU client was starting with windows.....so it tried opening multiple copies of the GPU client, which caused the error there.

edit, i just looked at core temp and saw something interesting.....no core has more than a 70% load going on it......could HFM have something to do with this? Yesterday when I ran this, before the new unit I was sent, when I watched it was doing the normal load the cores to 100%...

no hfm couldnt cause that.. check and make sure the cores set to run on all 6 threads.. 70 across all cores sounds like something is throttling it..

 

[W.Feather]

where do I go to check that its set to do that?

ive got nothing else but Chrome and the GPU client running and actively doing anything....the same as ive always had.

and its not 70 on all cores....mostly 40 on all cores, a single core will jump to 70 on occasion

 

[404 - File not found]

triple check your log file for the -smp flag when you started the client. Make sure you start the client by using the Shortcut you created with the -smp flag, clicking the client directly will not work.

Also monitor task manager by CPU rate, maybe a windows service crashed and is taking ressources? altought that's unlikely.
Posted via [H] Mobile Device

 

[W.Feather]

your 2nd suggestion was what fixed it...i was opening from the window HFM opened, forgot that I had to set an -smp flag for the shortcut.....thanks 404

 

[Nathan_P]

Ive got a AMD Phenom II X6 1090T(clocked at 3.2GHZ till it cools down), a PNY 260 c216(clocked at 576/1242/999), and am getting 12k ppd for both......is this about right?

I had it run for 3 weeks straight, and got 12k ppd on average for those three weeks, i dont know if that is about where I should be, I woulda thought that my x6 would get more than 10k ppd(I assume a 260 gets more than 2k..), just wondering, do i need to let it run for a good deal longer to get a decent idea on the PPD I get?

My 1090T@3.2 gets 10k PPD while driving 3 gpu clients.
your gtx 260 should get ~7.5k-8k PPD average

both these assume 24/7 running so that would be about 18k PPD

 

[W.Feather]

i get close to 24 hours a day, will be soon all week.....when classes start ill be in linux all the time(on the week days), so no games or CPU intensive stuff lol