So as per Geekbench, 15w i5 is faster than 28w i5?

[meboy]

The nTB 15w 2.3 with boost to 3.6 (I5-7360U) is faster than the TB 28w i5 with 3.1 with boost to 3.5 (I5-7267U)
Really?

Sounds good to me as I have zero interest in that touch bar, but a little hard to believe.
I tend to trust them for the most part though. They seem to be pretty accurate.
Maybe the 15w has much lower temps and doesn't throttle as much as the 28w. Maybe the cooling isn't good enough.

Thoughts?

https://browser.primatelabs.com/mac-benchmarks

 

[vatter69]

I own the 2016 ntB with the 2.4 i and i have yet to see the CPU throttle. So all this "but it will throttle sooner!!" is yet to be confirmed in 7+ months usage.

Even if you render a 2h 1080 movie in handbrake, no throttling. It get's warm and a bit louder , though.

 

[winterny]

Intel sells many different models of their various chips, but in reality they do not produce many different models ... They have a few different chips -- and once they are produced en masse, they are tested and undergo a process called "binning". The best performing chips are marked as the highest 'model', and chips which have lower performance are marked as lower models.

As the production cycle matures, a higher proportion of chips are produced which meet the higher requirements, and end up being marked as the 'lower' version simply to satisfy different pricing markets.

In short, some 7267U may be faster than some 7360U, some 7360U may be faster than 7267U. The minimum performance you are going to get for a 7267U is guaranteed to be higher than the minimum performance for a 7360U.

In reality, intel's processes are very mature, and most of their chips perform better than their base speed.