Swifty kit reviewed @ TomsHardware
 

Here is the review, of several kits actually.
I learned alot from TomsHardware in the past (nothing about WC however), and am interested how their review fares through the review crunching machine that is ProCooling as of recently.
( I dont have high hopes actually)
Right off the bat:
Though they summarize their tesbed to a good degree, there is no mention of where the temps are coming from, so one is left to assume that they are recorded from the likes of mbm5 or the BIOS.
Secondly, how and where they measured the dba ratings is a mistery.
Also they most likely did not do several mountings, of the blocks.

Anyways just figured you guys would be interested.
enjoy, hehe.

SenC.

What they hell did they do here?
On this page: http://www.tomshardware.com/cpu/2004...ooling-24.html

They got all kinds of shit hooked up to that. pH? Groth?

why do I feel like the whole toms hardware reading the CPU current thing was covered already?

but I also feel like it was unresolved

If it was I missed it. Been on the road a lot lately. Any link to a thread?

Curious. The Asetek Antarctica performed better with 7V fannage than with 12V fannage. Actually for a number of kits this happened.

Says a lot for the accuracy of the P4's thermal probe really.

How did that Titan thing do better than the Swifty or Asetek kit? :shrug: Tom's has been known to get paid off before, I don't trust them at all.

Well, that could be (probably wasn't, but it could have been) specific to that test bed. I find larger issues with the increase in performance with the lowering in fan speed. That just throws any legitimacy it had out the window, and this just shows that a P4's diode can't even be used for comparison.

This would be correct for P4's WITH the IHS on, with the IHS off, relative diffs are pretty good. However, all ASUS intel boards do not read on die, they read insocket, makes me wonder why they use this board for tests? Abit is the ONLY one im aware of that uses the ondie sensor.

The larger one looks like it displays the Vcore. Im not sure though, I didnt think P4s run @ 1.5v ??!!
The smaller meter , ( I think it reads, 043.2) temp monitor perhaps??
I'd really like to know, since they say they use that testbed in every test involving Intel.

SenC.

The doohickey in the back, with the three red wires running through it, is an inductive ammeter. 43.2 amperes.

Inductive ammeters are pretty good with AC, less so with DC, even less for DC with a bunch of AC on top of it. Plus, the way they added those extension wires to the torroids makes for tons of EMI.

Overall, their power measurement should be within ~3 percent.

on the right track....

overvolted P4s run that much, sometimes 1.7 (1.7intel is like 2.2AMD)

The larger meter is a DMM, measuring vcore

The smaller meter is a magnetic-based coil clamp on ammeter, measuring amps

watts=amps*volts

about 65w in this case

not totally absolutely accurate for measuring heat output, but good for relative measurements

realize not all the heat comes from vcore, not all the vcore turns into heat

dang groth beat me

Yay for that typing class I took! :p

nvm, only mentioned briefly when they were brainstorming for ideas to measuring heat output........like, last year

I would add that even with the Abit boards, the sensor doesn't seem effective with water cooling. With the amd boards, they use the sensor under the socket. Atleast with my NF7-S. With intel I couldn't say.

Almost, I can run 1.7v stable aircooled with my P4.