ECE.
Let's do a more in-depth look at the chassis, for the current discussion:
Claim (1):
The internal volume of air residing inside the 17" MBP is greater than the 15" MBP to the extent it will have significant difference in heat dissipation.
The heatpipe will indeed dissipate heat into the surrounding air via convection. Does the 17" MBP have significantly more air surrounding the heatpipe? I argue it does not. Here is an image (from iFixIt) of the 15" Late 2011 Macbook Pro heatpipe system:
Note it is the same design as the Late 2011 17" MBP I show in my video.
The 17" Unibody MBP is listed as being 0.1cm taller than the 15" Unibody. The length of hinge exhaust area is ~280mm on my Late 2011 17" MBP (would someone mind grabbing the measurements from the 2011 15"?), so you could make the argument that the 17" (assuming the 15" has the same 280mm exhaust length) has an extra 280 cubic mm of air. Approximately 0.02 cubic inches.
Ok, so assuming 25C ambient air temperature, specific heat capacity will be ~1005 J/Kg, with a density of ~1.18 kg/cubic meter.
How many kg of air do we have in 280 cubic mm? 2.8 x 10^(-4) kg.
How many Joules of energy can that air hold? 1005/1 = x/0.00028
x = 0.28 Joules. Are we really saying that the ability of the air surrounding the heatsink area to hold another 1/4 Joule is significant?
(to others following the discussion, 1 watt = 1 Joule / second. Therefore, a 45watt CPU will pump out 45 Joules/second, so the 0.28 J is ~0.5%)
Just to ensure clarity: I'm
not stating that design protocol dictates ignoring air volume,
nor that air volume means nothing. I'm only demonstrating that the air volume difference around the thermal cooling assembly between the 17" and 15" is
insignificant.
Claim 2:
Transistor switches do not short to ground.
I can't even begin to believe this one, I'm afraid we'll need to discuss it further.
The transistors build a CMOS gate in the CPU/GPU switch between VDD (positive) and VSS (ground). I'm *really* confused, are you claiming this untrue?
I think you may want to look up how a CMOS gate works. There is a definite rising and falling edge, and as they overlap, you have a short, and power dissipation.
Claim 3:
You can't compare different CPU designs and/or process steps at the same clock speed because they are doing a different amount of work for a given clock cycle
Sure you can, you can compare any two or more items in the universe. Whether anything
meaningful comes of that comparison is another story, however.
Intel is widely known to use their 'Tick-Tock' strategy for releasing CPUs.
You can compare C2D vs i7 for many things - Work/Energy used, for instance, and have a meaningful comparison there. (Tock vs Tock)
Tock vs Tick, though, is always enjoyable - per Intel, their Tick is simply a 'refining' and die shrink of the previous core.
Sandy Bridge vs Ivy Bridge, for instance, you can mark clear gains in power efficiency between the two, thanks to 22nm die + tri-gate design.
You can easily make the statement that Clock for Clock, Ivy Bridge should (and indeed, DOES) run cooler than Sandy Bridge, despite both being marked in that 45 Watt thermal envelope (I wouldn't be surprised to find that 2.5Ghz SB truly is a 45 Watt CPU, while 2.5Ghz IB is a, say, 41 Watt CPU)
So I don't understand your stance on this one. Are you saying no meaningful data can be had from comparing two different architectures?
Claim 4:
The 17" chassis is physically larger (has more aluminum) so the chassis will dissipate the heat more efficiently
Yes, you can argue that quite effectively. There is significantly more aluminum to the sides of the hinge area vs the 15", yes.
However, comparing a 17" vs 17" clearly won't have that issue. So another person that drops in with a 17" Macbook Pro wanting to know if 98C-100C under load is normal, when demonstrably the 17" should be running closer to 85C-90C under load. The same for idling @ 60-70C, vs the 17" should be idling @ 30-35C. (assuming STP for both statements, clearly)
The investigation into why one 17" (with a higher clocked BTO CPU I might add) runs
significantly (17% average) cooler than others, is, I think, a mighty fine topic of discussion.
Ok, gotta get ready for work, wrap this up Doward
I think it's clear that the Macbook Pro line up has the capability to run cooler, and if Apple would implement just a bit more quality to the thermal system, it would perform admirably.
I do believe the (albeit limited) data points to there being a clear issue of Apple's Notebook Lineup running on marginal cooling systems.
No, I didn't get into the thermal conductivity of air (poor), and yes, I am simplifying quite a bit - the idea isn't to impress everyone with my particular knowledge, the idea is to help foster knowledge amongst all. I've always said that's the difference between being smart, and being intelligent. Something I feel entirely too many of my colleagues ignore.
If you can't explain it simply, you don't understand it well enough. - frequently attributed to Albert Einstein
