In Which we discuss the next folding at home competition

Associate
Joined
25 Aug 2004
Posts
500
Location
Chester, England
Mattus said:
Would multiplying points by % overclock do something similar? Not a brilliant idea but all I can think of ATM.

That could work, but your problem then is that we need CPU-Z clarifications etc to prove overclocks which takes time, effort, and may detract from the ease with which people can start the competition.
 
Soldato
Joined
16 Dec 2005
Posts
14,443
Location
Manchester
My PPD has gone up about 600PPD after I borged a part time E6400 on Friday with crappy WUs.

My full-time OCd 4400+ only manages 600PPD with the best 600 pointers.

SiriusB
 
Soldato
Joined
31 May 2006
Posts
7,564
Location
West London
divine_madness said:
Do we know anyone folding on a stock E6300 and a stock 3800+ to compare though? :p

That what the benchmarks say e6300=3800+
Only unclocked C2D I know off is a e6400, and it just beats my 805D on proformance so thats sounds about right.
But you can Oc the pants of them (as we all know) almost doubling proformance.

Lady 3 - [email protected] =5000pts (5000/4.4) 1136
Bloke 1 - 4800+ =3800pts (3800/4.8) 792
SB - 4400+ =3800pts (3800/4.4) 864
Bloke 2 - e6300 =3000pts (3000/3.8) 790
SS - 805D =2800pts (2800/2.66) 1052

*numbers guessed
 
Soldato
Joined
16 Dec 2005
Posts
14,443
Location
Manchester
Your benchies must be wrong then, not only does my real-life example suggest the benchies are wrong, the CPU test in this months CPC mag has benchies for Fah and the X2s are no where near the C2Ds.

Not having a go but so far E6300 = 3800+ doesn't stack up :D

SiriusB
 
Soldato
Joined
27 Sep 2004
Posts
9,229
Location
Nantwich
e6300 are not as quick as people make out to be from what i have tested, i made a work rig for my dad a few days ago, did a quick set of benchies. sp1.5, CB & multipi, safe to say i understand the only reason people get them is to oc, at stock they are nothing really special.

How about a quad core league, for qx, quadfather and dual dualcore system owners?

That would rake in the wu's/points :p
 
Soldato
Joined
16 Dec 2005
Posts
14,443
Location
Manchester
I suppose lol.

I was thinking about a couple of things just now and we ought to sort them out.

1: We should decide on the format asap. Shadow's is my favourite minus the "wallet wonder" league.

2: We should start thinking about prizes and who is going to handle them. Assuming of course we able to do this. Would be nice if a Don could let us know for sure whether we can or can't have them.

3: We need to decide who is going to run this competition. I know everyone has ideas and input, but when it gets right down to it we can't ALL run it. Too many cooks and all that. Once we have finalised a format we can work out what tasks need doing and take a vote or something.

SiriusB
 
Soldato
Joined
10 Apr 2004
Posts
13,491
E6400 is nowhere near 115W load, even overclocked!

Intels top TDP they say is 65 or 75, can't remember which.

Even then F@H isn't using that much.

My E6600 @ 3.535ghz is using 110W roughly
 
Soldato
Joined
31 May 2006
Posts
7,564
Location
West London
Concorde Rules said:
E6400 is nowhere near 115W load, even overclocked!

that's the whole system btw - mobo/ram/hdd & the cpu :p

CR - thanks for the real life info - most of the figs are from tom's hardware (that totally non-bias site) so I'd take them with a pinch of salt.
Benchmarch are in our very own sticky!

But that said - it's all illustrative.
 
Soldato
Joined
31 May 2006
Posts
7,564
Location
West London
SiriusB said:
So who is volunteering to find all the TDP info for every CPU, Mobo and RAM that is to be entered?

SiriusB

Mobo (by chipset groups) and ram (by voltage groups) at stock are much the same. So I'll do that.
Oc'ed stuff will use more watts but I'll just make a reasonable guess if I cant find fig's.

I'll tabulate (and simplify) the results for publication prior to the big day. And do a wattage totalling service to all participants.

What would really help is if you guys could measure how much Lecky your rigs use and supply me with the details. That way I can double-check web info.

For those without a voltmeter, use your electric meter (set rig to fold cpu@100%, turn off screen - note value - wait 10mins - note value - power down pc - turn off at plug - note value - wait 10 mins - note value )

Note: It's obvious I know but try not to change background electric load while testing!!! (Read a book/use a battery operated electrical device/wrap presents/talk to someone)

Edit: for thoses with GPU folding - please complete test with only your cpu folding at 100%
 
Last edited:
Soldato
OP
Joined
18 Jan 2005
Posts
4,171
Location
Northants
How would one measure the pc power with a multimeter?

And using the main electricity meter is going to be difficult because you have to watch out for things like fridges starting up, heating and heaters coming on and off and whatnot.


I still can't decide which competition method i like best. I'm finding it a bit overwhelming the amount of good suggestions we have. I need to have a good long think.
 
Last edited:
Man of Honour
Soldato
Joined
2 Aug 2005
Posts
8,721
Location
Cleveland, Ohio, USA
Joe42 said:
How would one measure the pc power with a multimeter?
You would first measure the mains voltage in parallel with the load. This could just mean sticking the leads into the socket. You'd then splice into the power cable to measure the current in series with the load while the computer is switched on. You'd multiply the two numbers together to get the power because P = I*V.

For example if you measured 220 Volts and .5 Amperes you'd get 220 V * .5 A = 110 W.
 
Soldato
OP
Joined
18 Jan 2005
Posts
4,171
Location
Northants
BillytheImpaler said:
You would first measure the mains voltage in parallel with the load. This could just mean sticking the leads into the socket. You'd then splice into the power cable to measure the current in series with the load while the computer is switched on. You'd multiply the two numbers together to get the power because P = I*V.
Not exactly trouble free then. ;)
I don't fancy sticking a metre into anything mains powered and i don't fancy splicing any cables. I would be happy to buy a little power meter though and do some measurements.

If SS wants some power figures he should make his way over to some of the review sites which measure system power consumption of everything they test. Here for example.
 
Back
Top Bottom