• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

F@H Annual Holiday race ???

Markfw

Markfw

Moderator Emeritus, Elite Member
I have not seen anything yet. I will be busy next week, and out of town thanksgiving week, but I want to participate. Put me in if somebody starts it up.
 
Yes, there will be a race. But it will not start until mid-december.
Have patience --- there will be info in due time ... 🙂
 
Oh dear lord. Is it this time already? I said I would race this year.... I need to tear down my computers and REBUILD THEM just so I can F@H race this year? Again? I'm never prepared and my watercooled systems DON'T EVEN HAVE WATER IN THEM.

I need to buckle down over Thanksgiving and get them ready to F@H (except they flip the breaker in the computer room, so I'll have to move some elsewhere). And then I need to borg some work computers... we have a few Mac Pros and a few Mac Minis and then I need to set up GPU clients too.... so much work... so little time! Where's that CUDA hack for Linux?!?!?!
 
So a quick question... I haven't done F@H in almost 2 years....

If I'm running 13 SMP clients (5 Core i7's and 2 Core2Quads) and 4 GPU clients (2x GTX 260's and 2x 8800GT's), how many PPD do you think I could achieve and how much power would I use? I know it's different for everyone, but I'm estimating a power draw of 2kW continuous, would would be quite the power bill.

How much do you pay for power, if you don't mind me asking, ball-park, Mark?
 
PCTC2 said:
Where's that CUDA hack for Linux?!?!?!
Click to expand...

The instructions for running FAH/CUDA on linux are here

Notes of caution: 1) You need to use nvidia driver 180.60 or earlier and 2) kernel versions 2.6.30 and later are "broken" for FAH. If you are running ubuntu 64-bit or earlier, you should be ok.

I was hope there would be native support for FAH GPU2 by now. 🙄
 
Last edited:
PCTC2 said:
So a quick question... I haven't done F@H in almost 2 years....

If I'm running 13 SMP clients (5 Core i7's and 2 Core2Quads) and 4 GPU clients (2x GTX 260's and 2x 8800GT's), how many PPD do you think I could achieve and how much power would I use? I know it's different for everyone, but I'm estimating a power draw of 2kW continuous, would would be quite the power bill.

How much do you pay for power, if you don't mind me asking, ball-park, Mark?
Click to expand...

I get about 2-2.2k for my 2 I7 clients (@ 3.8) and about 1500-1700 for my C2Q clients (@3.4). As for the video clients, the 260s get 5k-7k depending on the core, and the 8800's will get 3-5k depending on client and how fast a card they are.

So, 33k total approx ???

Right now I have 12 video clients running and 12 CPU clients running. Not sure on this months bill, as last months still had a little AC on it, but it was $215, and I expect this one to be under 200.
 
Last edited:
Markfw900 said:
I get about 2-2.2k for my 2 I7 clients (@ 3.8) and about 1500-1700 for my C2Q clients (@3.4). As for the video clients, the 260s get 5k-7k depending on the core, and the 8800's will get 3-5k depending on client and how fast a card they are.

So, 33k total approx ???

Right now I have 12 video clients running and 12 CPU clients running. Not sure on this months bill, as last months still had a little AC on it, but it was $215, and I expect this one to be under 200.
Click to expand...

Thanks Mark. Looks like I'll be going all out this year for the F@H race, now that I have enough money for a month of excessive electricity usage. See you guys for the race!
 
Markfw900 said:
Why mid-dec ? we used to always start dec 1.
Click to expand...

Quite simple: I do not have the time to get it all fixed because I'll be out of the country for three weeks and I'll not have the time to set it all up before that.
I'll give it a try --- but I won't promise ... 😉
 
Don't worry Peter, the race can wait until you get back.

That might be better, it would split the electricity bill over two months for those concerned about that 🙂
 
I agree with Gleem.🙂 Might not be a bad idea. What the hell, give it a try and let's see if works out better by splitting the months.😎
 
dajeepster said:
Am i wrong for wanting F@H to get on the boinc bandwagon?... i would fold soo much more if they were
Click to expand...

No, not wrong, but I don't think it's going to happen.

When BOINC first started, F @ H had looked at using it, I think they even had a beta client.
 
GLeeM said:
No, not wrong, but I don't think it's going to happen.

When BOINC first started, F @ H had looked at using it, I think they even had a beta client.
Click to expand...

I agree, it won't happen. I think they want to keep their "independence".

To their credit, they have pushed alternative platform computing like GPUs and PS3's. I'm a bit frustrated that they don't have a gpu client for linux yet. Some very creative linux users developed a WINE wrapper so they could fold on their nvidia GPUs but now the latest driver version they can use is 180.60.

They say they will develop a linux GPU client once they work out the bugs in the Windows version. IMHO, that means it will never happen.🙄
 
Last edited:
Markfw900 said:
You need a 64 bit OS, and at least 3.6 gig free memory for that. I am working on getting Win7 64 bit installed.
Click to expand...
I'm a little confused here, let's see if I can get this right:

Straight Linux can run the smaller SMP WUs and get about 10K ppd - the memory usage is not lots.

Straight Linux can run the -bigadv SMP WUs and get about 20-25K ppd on an OCed i7 or better - the memory usage IS LOTS.

To run VMWare Player you need a CPU that has virtualization.

On Windows you can run Player with a Linux VM and do a little less than on straight Linux.

The -bigadv WUs would need 64bit OS because the WUs need at least 4.6GB?

Does VMWare Player need 64bit? In order to run the smaller Linux WUs?

.
 
GLeeM said:
I'm a little confused here, let's see if I can get this right:

Straight Linux can run the smaller SMP WUs and get about 10K ppd - the memory usage is not lots.

Straight Linux can run the -bigadv SMP WUs and get about 20-25K ppd on an OCed i7 or better - the memory usage IS LOTS.

To run VMWare Player you need a CPU that has virtualization.

On Windows you can run Player with a Linux VM and do a little less than on straight Linux.

The -bigadv WUs would need 64bit OS because the WUs need at least 4.6GB?

Does VMWare Player need 64bit? In order to run the smaller Linux WUs?

.
Click to expand...

You need 64-bit linux to run the client with the -smp option (big or small!)

Oddly enough, some of their code is also 32-bit so you also need 32-bit libraries too. 🙄

If you go with 64-ubuntu, you can install the 32-bit libraries with:

sudo apt-get install ia32-libs

🙂
 
Sooooo. I"ve DL'ed the winblows client and got it running........I take it, it does not automajically us both cores on a dual core processor 🙁. So do I need to run another instance of the client and set affinity? (Ahhh the old days of seti.)

I'll be along to pest ya'll with other ?'s as I try to set up other machines
 
GLeeM said:
What kind of dual core do you have?

biodoc, could Orange Kid run a "notFred"?
Click to expand...

Yes, I think he could run "notfred" on a dual core. Running smp on a dual core should produce more ppd than 2 x standard clients, right?
 
It's a pentium dual core T2080 The task manager is showing only 50% usage on the cpu.
tried putting another instance in a different folder but it whined that I was already running it once. I'm using Folding@home-Win32-x86. Where would I find "notFred"?
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top