Bench marking is one thing but sustained performance is another. Would love to see how their cooling solution copes with 100% cpu load for extended periods of time.
I've been testing the throttling on my Macbook Pro 15'' 2016 all weekend, by looking at Intel Power Gadget and doing various CPU intensive tasks. I was trying to decide if going for the higher CPU tier on the 2018 version would offer any performance improvements in real world apps or it would just be throttled down.
What I found.. well, it's not that great. My CPU is 2.7GHz, Turbo Boost to 3.6. In practice it never reaches that 3.6, best I got was 3.2. That's only if you stress a single core though, plus the dirty secret - in order to reach 3.2 on that core it drops the frequency on all the other cores.
If you go 100% load and stress all cores equally, it goes to 100 degrees Celsius really fast. It cannot maintain even the nominal 2.7GHz speed on all cores. Instead it goes down to 1.6GHz and a temp of around 80º C.
And it doesn't even have to be for an extended period of time. Just opening up Chrome with let's say 20 tabs, nothing too crazy, and let's say a video in one of them, would make it go to 100º C almost instantly and then it would correct itself by dropping the frequency all the way down.
It's not a surprise really... I mean, have you seen the cooling system for these things? It looks
like this
So unless they've substantially improved the cooling system in the 2018 15 inch models, going for the higher tier is probably a waste of money. I mean, it will be slightly faster just due to the bigger 12MB level 3 cache, but I'm not sure how much of an impact that would make. All I know is: 4.8 Turbo? no freakin way