Chimp Challenge 2009 thread

[jws2346]

Yup,yup, yup As I posted in my previous thread I was doin' the "test the new M$ Win 7 beta OS" on my stock city E6600 (past tense, I'm waitin' for a RTM and a few SP's now ) . Welp I reformatted and reinstalled good ole' LinSMP under the Ubuntu v9.04 jaunty jackbutt OS (or whatever it is ) and my ppd went from about 1300 ppd (the best) to ATM 1960 ppd on a WU 2675. (love dese 2a cores )

Now I dern't know about you boys and girls, but that's a hell of a ppd boost to me. Next to the GPU2 client, if at all possible, using the LinSMP is a no brainer. I realize for different situations and reasons using Linux for LinSMP may not be remotely possible or even feasible in some circumstances. (without the VMWare jobber doober )

I wish I could GPU2 under Linux (not the VM crap, I'm talkin' straight ) I like this Ubuntu v9.04 IMHO with the a2 core and a stand alone GPU2 client for Linux we'd haffta' raise the Chimp Challenge to 50 million pernts a day or 75 million pernts a day

DCing for the CURE

 

[swoody]

I wish I could GPU2 under Linux (not the VM crap, I'm talkin' straight

That would be awesome!! My GPU2 client is the only thing stopping me from being completely Windoze-free Everything else I own is already Ubuntu-only

 

[Xilikon]

Bumpppp

 

[tjmagneto]

My rig is ready to go for my first Chimp Challenge. Getting about 12K ppd, give or take a few with my i7-920 running at stock with a new BFG 275 OC, Windows SMP clients x2, GPU2 client.

 

[Kendrak]

We got some front page pimpage. I'm going to send a PM that it is to 20 mill and not 5.

 

[APOLLO]

We got some front page pimpage. .

Thank the gods...

 

[swoody]

Woot! I have a 9800GX2 coming just in time for the challenge! I'll pitch that in with my Phenom II for some PPD

I'm really glad to hear the goal has been lifted. 5mil would have taken a couple days at most

 

[Modred189]

My GTX 260 and ATi 4650 are in.

 

[84gthatch]

sweet keep it up everyone!!
i started switching my vm's over last night with a couple gpu's tonight i'll probably finalize the switch so in case tuesday gets crazy i don't miss a day.

 

[Modred189]

By the way, is it possible to run CPu AND GPU folding concurrently?

 

[84gthatch]

modred absolutely if you need help getting it set up just ask

 

[Kowan]

Yes you can run the CPU and GPU clients on the same system.
I run a SMP and two GPU2 clients on my second box.

 

[Xilikon]

To have a good start, we need to switch clients today

 

[Kendrak]

To have a good start, we need to switch clients today

We need a rolling start to avoid what happened last year.

 

[digital_exhaust]

I plan on having my boxens up and folding for the Ape by tonight... should be able to get things set up after work...

Beers and tamales guys.... let's get going!!!

 

[sirmonkey1985]

By the way, is it possible to run CPu AND GPU folding concurrently?

i run 2 SMP clients in VM + a gpu2 client..

if you have an ATI card you have to leave a cpu core dedicated to the each gpu client.. who knows maybe one day panda group will pull their heads out of their.. and fix the gpu client on the ati side..

 

[jws2346]

Welp, I just switched my little folding garden over to [H]ardApe (I'm gonna' help keep some sneaky snake from gettin' the jump on us this year)

I kept my GPU2 clients runnin'. I shut my WinSMPs and LinSMPs down completely (even shut down my Linux boxen to save some electrcity) Is FAHmon pretty accurate when it comes to when the WU was downloaded and when the WU's preffered deadline is

I ain't suggestin' youse' switch over this early. I just couldn't help it (the foldin' devil made me do it) To repeat an old phrase used by a past US President I say "bring em' on" to anyone that would challenge the mighty [H]orde

We's not some of the best, WE are the BEST

DCing for the CURE

 

[digital_exhaust]

Um.... I hope I'm wrong, but.....

Just looked at the stats page Xil linked in the OP and MaxPC, OCN and our friends down under are all on the top producers list... and we aren't....

Is this a bad thing, or does it not really matter yet??

AND... when searching for [H]ardApe, there are several entries showing up. It seems the second listing down on the list is the active one... I guess I'm just concerned that maybe there are some that have already turned their boxen over and perhaps have entered the wrong name...

Thoughts??

 

[Xilikon]

Anything can happen during a challenge so it's best to dive into it and pray for the best It's like a F1 racing... Anything can happen

 

[TheMead]

I just upgraded and think that folding might be a great way to stress test my hardware.

To be honest I haven't folded since the last challenge and I haven't keep up with the in's and out's of folding yet. What would be the best client for me to run?

My Hardware: AthlonX2 7750 BE (kuma), 4G ram, 9600GT currently running XP SP3.

Could I run both a gpu based client and a cpu client since I have a dual core cpu?

 

[swoody]

To have a good start, we need to switch clients today

Got my box switched over. Not going to make a lot of PPD until I get my 9800GX2 in tomorrow

The Mead - Yup, you can definatly run both a CPU and GPU client. I'm not sure how much CPU power that GPU client will use in XP, but you could either setup an SMP client to use both cores, or if the GPU client is hording a bit of your CPU, you may be better off running a regular client on your CPU. Hopefully someone more knowledgeable on this will cut in here

 

[84gthatch]

^^ i'd take alook at the vmware thread and get one vmlinux setup and then one gpu client setup and that would be a good place to start.

 

[jws2346]

Err..., I don't want anyone to think I was trying to cheat in any way. AFAIK any pernts turned in before the May 5th 12 PM accredited to [H]ardApe DO NOT count toward the Chimp Challenge total.

What I was doing was making sure my boxen were switched over (albeit a little on the early side). Also from last year, I was a little miffed by the fact, compared to the other competitors, we ([H]ardApe) were in the freakin' hole from the jump. Now I ain't sayin' anyone held off from returning WU's, but it was suggested it might have been done last year . If it turns out I did something this year that I heard about too late last year I'll be sure to cry my eyes out and sacrifice a few chickens to the folding Gods

Please don't anyone get their panties in a twist, these are only my lowly opinions and actions, not suggestions)

DCing for the CURE

 

[digital_exhaust]

I don't think anyone is thinking you were cheating.... it's all good man.

Unless I'm misunderstanding something, it's in our best interest to start early, that way we are already rolling one the whole thing officially starts, right?

Anyhow, fold on man!!

 

[Modred189]

Ok, so I was thinking, you can run F@h in linux. I have virtual PC installed on my system with an Ubuntu system set up.
I installed it (following their directions from the installation page, because I hate the command line), and it works. My question is whether or not that will result in any kind of real numbers compared to any other option out there (that I dont' know about).

More importantly, is it possible that running the virtual PC is stealing enough cpu cycles from the GPU folder to actually decrease the total folding rate?

 

[Axdrenalin]

Ok, so I was thinking, you can run F@h in linux. I have virtual PC installed on my system with an Ubuntu system set up.
I installed it (following their directions from the installation page, because I hate the command line), and it works. My question is whether or not that will result in any kind of real numbers compared to any other option out there (that I dont' know about).

More importantly, is it possible that running the virtual PC is stealing enough cpu cycles from the GPU folder to actually decrease the total folding rate?

I was running two VM's under VMWare server, and the GPU client on a Windows XP Quad 6600 system at 3Ghz. I noticed that when I ran two VM's the GPU took a little longer to complete a WU. I dropped it down to one VM and the GPU processing worked out better as well. I'm assuming based on these results that the GPU performance is tied into the CPU somewhat.

What kind of CPU are you running your system on, and how many cores do you have tied to your VM / VB? NVRMIND - if you running on your sig setup you should be good to go...

 

[Kendrak]

GPU needs 1 or 2% to feed the client with a NV card. doesn't need much, but it needs to be ther or the client will starve.

 

[Modred189]

I was running two VM's under VMWare server, and the GPU client on a Windows XP Quad 6600 system at 3Ghz. I noticed that when I ran two VM's the GPU took a little longer to complete a WU. I dropped it down to one VM and the GPU processing worked out better as well. I'm assuming based on these results that the GPU performance is tied into the CPU somewhat.

What kind of CPU are you running your system on, and how many cores do you have tied to your VM / VB? NVRMIND - if you running on your sig setup you should be good to go...

It is on the setup in m y sig. Thanks.

 

[rmdashrrootsplat]

Got the Q6600 at work crunching away at two VM's... Will swap over my home pc tonight.

Hell, I just may fire up the old P3 lappy and run the uniproc on it. It just may be the grain of rice that tips the scale for us!

 

[lassiterb]

Starting to switch all of my farm over .......................................... now!

Got a few borgs that I won't be able to switch over but all that I have under my control shall join the CC.

Also I have parts arriving on May 6th for a quad GTX275 box, hopefully I will be able to get it put together and running before the CC is over.

 

[Rebel44]

All my 3 PCs are switched to [H]ardApe

 

[84gthatch]

got about half my stuff switched over will finish the rest up tonight we need to be rolling 1 million ppd at least by tomorrow probably alot more!

 

[swoody]

Ok, so I was thinking, you can run F@h in linux. I have virtual PC installed on my system with an Ubuntu system set up.
I installed it (following their directions from the installation page, because I hate the command line), and it works. My question is whether or not that will result in any kind of real numbers compared to any other option out there (that I dont' know about).

More importantly, is it possible that running the virtual PC is stealing enough cpu cycles from the GPU folder to actually decrease the total folding rate?

Well, for me, it depended on the priority that I had my GPU client running at compared to the VM's. I now have 2 VMs setup at "AboveAverage" Priority, and my GPU client setup at 'RealTime' priority. As long as your GPU clients are setup above the VM's you should be alright, as it'll give your GPU 'first dibs' on your CPU power, and then pass the rest along to the VM's. WinAFC can help you out in setting up priorites and such for all your clients:
http://affinitychanger.sourceforge.net/

 

[ALL4AMD]

my 4850s and laptop will be in

 

[knob]

Main box churning for the Ape, htpc churning for the Ape. I'll pause all my WUs near 99% so I can package/upload completed units just after the opening bell rings.

Unfortunately, my GTX260(216) ain't here yet from the 'egg (deliver tomorrow)... so I'll have some downtime tomorrow at the beginning of the race. It'll get installed after I get home from work, and teh 9600GT bumped to the htpc shortly thereafter.

Exit question: does PPD scale favorably with overclocking on a GTX260? if so, which clock(s): core, shader, vRAM?

 

[84gthatch]

shaders are the primary oc point for good ppd and yes it does scale relatively well

 

[Air Cooled 420]

Scales well adjusting shader clocks.

 

[Air Cooled 420]

Broke down and ordered an I7 for CC, or at least that is how I will justify the purchase for now For The Horde!

 

[84gthatch]

Broke down and ordered an I7 for CC, or at least that is how I will justify the purchase for now For The Horde!

lol whatever it takes bud I about ordered 4 260's and a new rig to go with them lol -_-

 

[rcranfield]

Best of luck to you guys in the CC09. It should be both fun and interesting. It is our first time in this over at EVGA.
I guess it is time for us to put up or shut up?