Commander Kinsey wrote:
All product specs are lies. The CPU cooler in this computer claims it
will cool a 300W CPU. Mine gives off only 140W, and that gives a
temperature rise of 55C, according to both my readings and those of 3
review sites. 300W would cause a temperature difference of 120C, so to
keep it below the usual maximum of 100C, you'd need a room temperature
of -20.

Even the 140W is odd, this CPU is rated at 105W. Maybe they don't
account for you using every single part of it simultaneously.

And how do you know it's 140W ?

What would the error bars be on that measurement,
in your considered opinion ?

One site used to measure the coolers, and it
appears to still do some of them.

https://www.frostytech.com/articles/2766/4.html

Taking an example:

Noctua NH-D15 high 10.9C for 125W input or 0.087 C/W.

If room is 25C, inside case air is 35C, 140*0.087=12.18C rise or 47C.

And the processor in question, never gets even remotely
close to throttling temperature. If we try the maths at
300W, 300*0.087=26C and 35C case air gives ~61C at the plate.
There is a temperature rise to the CPU die (cooler plate to
CPU die path) not covered by that, but it's going to be
under 100C at the silicon die. It would pass by a little bit.

If we knew your cooler model, it would be easier to
work out examples. Maybe Frostytech never measured it,
as there's a lot of coolers out there. And measuring them
all with a constant jig setup, gives a chance for a fair
comparison.

Paul