I think the calculator is a bit out with regards to the Q6600 power usage, it reckons it should be using 130w at 1.25v when the official specification is 105w at up to 1.372v.
This means once you enter an overclocked figure the TDP goes up a lot, in reality it'll be considerably less than the calculators predicting.
I run a Q6600 at 3GHz and 1.4v and it reckons that is over 200w, which is frankly ridiculous.
I have emailed them, will post here what they say about it.
Personally, I dont think its as far off as you do, but we will see
You owe it to yourself to click here really.Originally Posted by Blitzen
I use this one rather than the "eXtreme" one, as I find it more accurate overall. As far as CPU draw is concerned I've no real idea, but my impression is that the eXtreme calc is generally overestimating wattage requirements, sometimes quite considerably.
That one looks a LOT more realistic
Ok, got an email back, they had the standard TDP down as 130w, thats why everything else was showing as higher.That one looks a LOT more realistic
Earlier on, I did;
3600mhz
1.6v
it gave me 319w
Now it gives me 258.
The other PSU calculator that was posted is flawed as well, in so much as it adds wattage when you add mhz, and not just voltage.
It shows 238.39 for the above settings.
Maybe somewhere in between as a happy medium eh?
You owe it to yourself to click here really.
Like I said, I've no real idea when it comes to CPU draw. However, I certainly trust "mine" more in regards to everything else, after having compared both against known power draws for various configurations. Not to mention that it gives up the 12V draw for free.... But yeah, for CPUs I'd likely go with somewhere in between too, just to be safe.
Errrrm, increasing the frequency will increase power draw.
I believe its roughly linear with clock speed and squares with voltage.
No harm done, its good to find out stuff like this
Last edited by Clunk; 04-06-2007 at 07:21 PM.
You owe it to yourself to click here really.
I think we have crossed wires.
It was showing the same wattage for the same overclock with or without 1.6v. I just retried it again and it seems to be ok.
Thorburn, are you using watercooling?
I could do with someone with a similar setup to mine to tell me what their load/idle temps are at stock volts/mhz, also room temps, and then maybe at 1.4v, 1.5v and 1.6v (if you are ok with doing that). measured by coretemp if possible
I have a feeling that mine is reporting temps a mile off, tried everything, can only be the cpu/board now.
Last edited by Clunk; 04-06-2007 at 07:57 PM.
You owe it to yourself to click here really.
Thanks all!
I've read your guide Clunk, went to some reviews for my ram, ps, read some more on wikipedia... i'm starting to "get it" (a bit ;-)
So uhm, in a push pull config should i put them blowing from top to bottom of case? front to back? Fastest should be pushing or pulling? Why not both pushing from top/front instead?
I guess I'll just try but if you have any idea why one is better then the other I'd be curious. My guess would be top to bottom since the PS in the P182 is pulling air there and i think the top fan is pushing it... (and my first guess would have been both pushing at 90° if i didn't read about push-pull in reviews)
Last edited by RRA; 04-06-2007 at 10:33 PM.
I'm not too sure on how well push pull works on that heatsink, you might get more answers if you start a thread about it
You owe it to yourself to click here really.
That was an interesting discussion regarding estimation of CPU power draw. Perhaps the accuracy of the calculators being discussed could be tested using some simple, rough experiment?
Since CPU power on currrent motherboards are almost exclusively drawn from the 4- or 8-pin 12V connector, we could either use a clamp meter or build a pass-through connector (for use with normal DMMs) to measure the current going through that connector. We can then use a DMM to measure the exact Vcore (positive probe on the drain tab of one of the VRM MOSFETs, negative to ground) or just use the BIOS/software readings. Multiply these two and we have a good estimate of CPU power draw. We then change the Vcore and/or FSB and repeat the test. Would that be sufficient for our purposes? Just a thought
Right i've got a really daft question, so here goes, please don't laugh!
Just got my P5K deluxe and corsair HX520 and i just want to no what connectors people are using to power the board? Do you use the EATX (the 8pin one) or just the single 12V ATX (4pin one) basically do i remove that little blanking plug or not cause i can't seem to find owt about it in the manual! Other thing is, is it dependant on what processor you are using?
You don't have to use the 8-pin, but you might as well.
Just pull the little blanker off.
It's almost certainly the 8 pin you should use..dare i say..read the manual!
Right i've got a really daft question, so here goes, please don't laugh!
Just got my P5K deluxe and corsair HX520 and i just want to no what connectors people are using to power the board? Do you use the EATX (the 8pin one) or just the single 12V ATX (4pin one) basically do i remove that little blanking plug or not cause i can't seem to find owt about it in the manual! Other thing is, is it dependant on what processor you are using?
Different board but similar problem...some boards wont boot up with out that supply, and I had the same problem, so fitted the four in (didn't remove blanking plate) and it works ok.
There are currently 1 users browsing this thread. (0 members and 1 guests)
Posting Permissions
LinkBack URL
About LinkBacks
Reply With Quote