MacBook Pro M2 Pro Review – Incrementally Better + In Depth Display Analysis

Apple’s M2 Pro MacBook Pro is great. It’s only a slight improvement over the last generation, but considering how huge of an upgrade that one was, it’s gonna be pretty hard to surpass that. Interestingly, not much has actually changed. Really, almost everything, save for the processor at the heart of this stylish anti-consumer machine is the same – so I guess we should focus on that.

Apple Silicon was arguably the biggest leap in outright performance, efficiency and control that Apple has made. Switching from Intel’s x86 chips meant Apple could introduce a significant number of new features and advantages – things like a hybrid core design with low power efficiency cores and high power high performance cores, impressively powerful graphics processing, built in video encoders for things like ProRES video editing, combined system and video memory that’s literally built onto the same package as the CPU itself, and even things like moving the SSD controller onto the chip such that the “SSD” that’s in here is actually just the NAND flash chips and wouldn’t work in any other machine. 

The biggest leap by far though was in efficiency. This entire machine only draws 52W while rendering in Blender. For some context, a lower power CPU like the 13700H in the Vision Pro 16 I reviewed recently peaked at 59W JUST FROM THE CPU doing the same render in Blender. A somewhat similar performance chip – the 12700H from Intel – draws 107W just from the CPU alone doing the same Blender render. Just the CPU draws double the power of the entire MacBook for basically the same performance. I’m still blown away by that. 

The performance itself hasn’t jumped up all that much from the M1 Pro. We do now have 12 cores – 8P 4E – but relatively little seems to have changed in the core architecture front. Cinebench R23 now nets around 1650 points in single threaded, up from 1533 on the M1 Pro, although that’s still behind basically every modern Intel or AMD laptop CPU, with the more performant ones pushing over 2000 points. In multithreaded the M2 Pro nets just shy of 15,000 points. That’s up from 12,335 on the M1 Pro, but again that’s behind basically everything from the 13th gen Intel lineup that I’ve tested, with the 13900HX hitting functionally double that performance. 

Blender shows a frankly incredible improvement, with the much longer Gooseberry render dropping from a touch under 19 minutes, to just over 13 minutes. That’s 30% faster! Now some of that is from the chip itself, although if my memory serves me right, Blender hadn’t released an Apple Silicon native version when I reviewed the M1 Pro, so some of this improvement I think comes from not needing Rosetta to translate the x86 instructions anymore. Still, it’s an impressive improvement. That still puts it in the middle of my test results, with the high end 13900HX running another 35% faster. Admittedly, that is drawing around 3x the power to do that though.

Just from these results you can see that the jump from M1 to M2 isn’t exactly massive. It’s a welcomed improvement of course, but to borrow Intel’s language, it’s very much a “tock” generation, when M1 was a massive “tick”. For your average user – hell even most power users – there isn’t going to be a world-changing amount of performance that you don’t currently have access to with an M1 chip. If you are just buying into the Mac market, then the M2 chip is pretty much all you can get your hands on now anyway, so you’ll at least be happy knowing it’s a decent chip.

So if that’s all that’s new, why is there still several minutes worth of this video left to run? Well back when I reviewed the M1 MacBook I hadn’t built this, the open source response time tool pro. I designed this thing to test basically any display, and while I still haven’t written a MacOS app for this, using Moonlight – the excellent game streaming software – I finally got the chance to properly test this MiniLED display. I think I might be the only person in the world to be enough of a nerd to have run this test, so let me show you what’s going on here.

Last time I used a high speed camera to show you the frankly horrific response times, and the really interesting but visually poor backlight effect this panel has, but now I can show you what’s actually happening with microsecond accuracy. Here’s a full black to full white transition. The first thing you might notice is, well, the noise. I’ll overlay what the “ideal” version of this transition would look like, basically a straight line that goes from functionally complete darkness to the upper target somewhere around 500 nits. Instead, thanks to the MiniLED backlight requiring a pulse-width-modulated backlight signal, you get this flicker. Zooming way in you can see what I mean. This is going from around 50% brightness to around 130% brightness, oscillating in around 20 microseconds. My best guess puts that at around 50,000Hz, which is more than enough to not bother even sensitive eyes. 

Let me add my denoising function to this so you can see a cleaner result. See, that looks much better! It even looks close to what I said the “ideal” result would look like – except I left out one little thing. That isn’t actually the “ideal” response. An ideal response is basically a straight vertical line. More like this. OLEDs functionally do this, but because MiniLED displays are just regular LCD panels with a fancy backlight, they are still susceptible to this sort of slow responsiveness. A good LCD will still only take a few milliseconds to do this sort of light level change, or at least take as long as the refresh rate window – as in how long between new frames. At 120Hz that’d be 8.3ms. This MacBook display? Yeahhhhh so it’s.. Erm.. 60.. 64.. 64 milliseconds. That’s EIGHT times slower than the refresh rate. Yeahhhhhhhhhhh.

Let me show you the heatmaps where you can see that this display is just.. Awful. Like, I’ve tested projectors that have faster response times. The lowest result is still slower than the refresh window at 8.8ms, but the average is an appalling 24 milliseconds. That’s three frames at 120Hz. Considering Apple put so much effort into their own CPU design, surely they can work out how to overdrive a display, right? This translates to a slow, jelly-like visual to any fast moving content, makes gaming even more impractical than it already is on MacOS and Apple Silicon, and is just visually off putting once you notice it – which I’m sorry to any Mac owners who now notice how bad their displays are. 

Happily the colour accuracy front is much, much better with my Datacolor SpyderX reporting 97% coverage of the DCI P3 spectrum and a DeltaE average of just 0.78 – a phenomenal result. It hits around 500 nits at full SDR brightness, and while I can’t test it, I completely believe it’ll hit 1500 nits in HDR. It’s blindingly bright in HDR. It’s just a shame that it has such a weird behaviour when it comes to response times – and that it takes so damn long.

Overall, my conclusions about the MacBook Pro haven’t changed much from the M1 Pro version. It still isn’t for me, and Apple’s constant anti-consumer behaviour is repugnant. A great example of that is demonstrated in a video by Hugh Jeffreys where he swaps the display on two identical M2 MacBook Pros and they break themselves. That means if you ever need to replace the display, you’ll need to not only buy the parts from Apple, but have them “reprogram” the display so it works as intended again. Naturally that service, plus the replacement display, will likely cost more than just buying another MacBook – and of course you can’t take your data with you as the SSD is now soldered to the board, and even if it wasn’t it’d just be the flash which means it’s garbled noise on literally any other machine, even an identical Mac. 

BUT – if you can put up with a company that actively despises you for not permanently emptying your wallet into theirs, I can’t deny it’s an incredible machine, both physically, and from a performance and efficiency perspective. Apple Silicon seems lightyears ahead of x86 chips, if not in outright performance, at least in efficiency, and for a portable device like this it’s hard to argue that isn’t an amazing benefit. I’m not going to be changing any minds here, so I’ll stop my ranting and raving and hand it over to you. What do you think about the M2 MacBook Pro? Let me know in the comments below.