Update: Disregard this, it's an unreliable piece of shit. I'm on my 4th one. 2017 version might be better.
Reviewed is the upper spec 15" 2016 stock model (i7 6820HQ, 16GB RAM, 512GB SSD, Radeon Pro 455)
Usual disclaimer: There is a lot of rambling here and I don't proof read. Expect lots of mistakes
Apple hasn't had a good year, the removal of the headphone jack on the iPhone 7 was perceived as user-hostile and entirely unnecessary (and I'd agree). Next up is the Macbook Pro, specifically the 15". Apple's absurd and totally indefensible decision to time-after-time ignore basic maintenance of their Macs resulted, in this case, a lot of pent-up demand.
The 15" MacBook Pro hadn't been properly updated since 2014, it received a good GPU boost with the swap to AMD's m370x from NVIDIA's 750m, but everything else remained the same. This wasn't necessarily a bad thing for many reasons, but given the AMD GPU wasn't particularly new in its own right and the other aging internals, enthusiasts and pros felt inclined to wait, even if CPU performance was unlikely to be any better.
However, as time went on, lots of peripheral upgrades came into view. Raw CPU performance these days can't necessarily be a motivation on this calibre of PC, we haven't seen any real improvements on that front in many years. So the focus turns to chipset functionality, specifically on the I/O front.
USB-C, USB 3.1 and Thunderbolt 3
Compared to previous generations, we have native USB-C, USB 3.1 and Thunderbolt 3 on Skylake. These are all pretty exciting developments, in particular USB-C/3.1 - Thunderbolt 3 is definitely more capable but less likely to be natively supported by peripherals, luckily one port can do it all. I use my 'development' laptops in 'docked' mode for a significant portion of my total use time, I want them to essentially be indistinguishable in as many ways as possible, I want multiple big screens, plenty of desktop like accessible I/O and the ability to extend with desktop like compatibility (via add-in cards or in this case fast PCI-E ports).
Amongst a few other things, these are the primary benefits of Skylake over Haswell, and at least for me, it's what I was waiting for. Now, was I waiting for Apple to remove all ports other than 4 Thunderbolt 3 ports? no, but of all the changes to the new MacBook Pro, this is the one I'm most on board with. Except the SD card slot, I don't use it, but I think that's probably going to be the biggest headache for many users.
The USB-C form factor and the capabilities it brings, even ignoring Thunderbolt 3, are pretty exciting. It brings a truly universal cable, it can be used for charging & data simultaneously, meaning for example a monitor could charge a laptop whilst being used as a monitor and acting as a hub in itself for other standard I/O. Thunderbolt 3 natively carries DisplayPort too, meaning capabilities such as daisy chaining or docks carrying multiple displays from one cable into the PC is possible without the display being Thunderbolt native, as was the case with previous generations.
This is a transition worth doing, there is some confusion and some annoying little side issues, but it is, I believe a great benefit for us to be moving towards this. Creating demand for the market by shifting popular devices to it is a good thing.
However, as usual, Apple is unnecessarily hostile in this transition. Do I think they should have compromised on the available bandwidth for USB-C / Thunderbolt 3 ports and included legacy USB-A ports? No. Do I think they should have included at least a couple of USB-A adapters in the box? Absolutely. They're inexpensive and it goes a long way on such an expensive device to ease the transition.
Not including such adapters shows a contempt for the customer that is becoming all too familiar with Apple. They don't care to explain or try and help you understand, they just do it and do little to help you adapt to these changes.
I'm not a fan of the MacBook keyboard (new thin and light model), whilst some of my favourite keyboards exist on both side of the key travel scale (lots of travel vs low travel), the MacBook keyboard dropped below what I felt was acceptable. It felt a lot like hitting a thin bit of plastic - it offered no feedback and I felt would have been fatiguing long term.
So, to say I was disappointed by the inevitability of this or an iteration of this keyboard making it to the MacBook Pro is an understatement. I was absolutely dreading it, I went into the 2016 MacBook Pro expecting to mostly dislike it as much as the MacBook.
However, to my surprise, the simple addition of a chunky and consistent click transformed the keyboard to something that has the potential to be one of the all time best laptop keyboards. The advantage of the wider, lower travel sturdier keys over the bigger chiclet keys of the previous generation MBPs (and similar laptops) is stability and uniformity.
They don't suffer that typical mushy and wobbly characteristic of those previous generations. That's an improvement that leads to quicker and more accurate typing with a nice satisfying feedback in the click (more so than those chiclet keys with longer travel).
Unfortunately, this is at the expense of mid to long term comfort. I'm typing this post out on the MBP keyboard and my hands are beginning to ache - this isn't typically the case with keys with longer travel.
To summarise on the keyboard
- Unusually good clicky feedback for a thin laptop keyboard
- Wide and stable keys for better accuracy and speed
- Lower profile allows for a thinner laptop
- Low travel isn't as good for comfort in longer typing sessions
- Arrow key design is stupid, it's not as easy for touch typists to be able to quickly distinguish between the up and down arrow keys and that's a problem.
I feel with maybe 25-30% more travel this could be a really fantastic keyboard, but the likelihood of us getting that seems slim, Apple isn't going to increase the thickness of the laptop for the keyboard of all things.
The new trackpad is up to 2x larger, it's massive and my initial reaction to the news was positive. Apple's trackpad hardware and software is the absolute best and the gestures it enables in macOS are a big reason for making macOS such a joy to use on the laptop.
More room for those gestures is surely only a good thing? That's what I thought anyway. Then people began to point out what should have been obvious, you're more likely to be resting your palms on the trackpad whilst typing - I worried about that - but had some confidence that Apple would not have designed this in such a way without having good palm rejection software.
Fast-forward to me actually getting the laptop and I was greeted with another surprise, I simply can't type in such a way that leaves my palms resting on the trackpad. I tried, I'm trying now as I write this, but the way I type on laptop keyboards at least has always had my hands arched into the keyboard, rather than resting flat on the laptop.
Is that to say everyone types like me? Absolutely not and there have been a fair amount of user reviews of people who do rest their hands in such a way that they rest on the trackpad whilst they type.
For those people, the palm rejection does seem to work most of the time, but that percentage needs to be higher. So whilst I do like the larger trackpad perhaps, given the general lack of additional utility afforded by it, making it quite so large was not necessary given the potential trade offs for some users.
A MacBook hallmark, this is something no one expected to be anything less than great. It is great. It's superbly finished in the new space grey, it fits together perfect, it feels good and the little details are all still in tact.
Details such as a correctly weighted hinge is something that many manufacturers fail on but Apple has nailed time and time again. On the topic of the hinge, it is now part of that 'unibody' design, entirely metal with no plastic inserts or components. A nice aesthetic upgrade.
Cooling is just as good*, the intakes are precise and have an oddly pleasing aesthetic value and outlets in the hinge are discrete. The laptop is quieter than ever too, you can make them audible in very intensive tasks, but even in longer term moderately intensive tasks it's inaudible in standard situations.
*By cooling I'm describing largely describing the acoustics. Unfortunately Apple has long shown little regard for running CPUs and GPUs are very, very high temperatures - but it has mostly worked well for them so far - and they're still within the limits of the chip designs, so I suppose it isn't something we should worry too much about.
The weight saving, on paper, should probably be more noticeable but I have found the reduction in overall volume to have the greater impact. It's not a huge improvement, and there is plenty of argument of leaving the volume of the device as it was with the Retina MacBook Pros to retain battery capacity, but that is seemingly an argument that manufacturers such as Apple aren't interesting in hearing.
One of the more surprisingly controversial things that I didn't think about at all until people began to ask me about, was the replacement of the glowing Apple logo on the lid of the laptop. I've been asked several times by several people if I miss that and the truth is.. I cared so little for it that it didn't enter my mind. I didn't think it was a good design element, especially not in recent years - it was a 'look at me!' tacky characteristic of the Apple machines but one I got used to.
The new logo is one that I think fits with the design of the newer unibody MacBook Pros (going back years) better. I do have some concerns about the mirror/chrome-like finish of the new logo, however. I'm not expecting it to age well once it gets a few scratches on it. And it will get scratched.
Either way, as I said, I can't believe people cared that much about it. Even despite not really being a fan of the glowing logo, if it was on this iteration of the laptop, I'd have probably not given it a thought other than the missed opportunity to make the display construction thinner.
It's great, I have some other excellent displays on many other devices and in terms of colour this one stands out. It's the same 2880x1800 resolution as before.
There really isn't much to say about it, it's an excellent well calibrated display and the wider colour gamut is noticeable especially when comparing it to other monitors.
The improvement to the speakers on these laptops is pretty impressive, the sound is good quality, detailed with a good level of bass. As someone who doesn't have any external speakers hooked up to my desktop PC, my laptop has always been my go to for when I don't want to wear headphones to watch a video or listen to a bit of music in the background.
So the improvements here are actually quite impactful for me and very much appreciated.
TouchBar and TouchID
The headlining feature of the new MacBook Pro, it has attracted a fair amount of discussion. It's not why I bought the laptop, and I don't think I'd miss it that much if it suddenly disappeared but I do think it could have some cool utility in the future. The most use I get out of it at the moment are in two ways:
Universal web media controls via Safari.
I listen to a lot of podcasts on the web (via pocketcasts), and being able to scrub through, play & pause podcasts even without the browser window or tab being in focus is very cool. These media controls seem to work across all standard media players (e.g. not ones that requires a plugin like flash or silverlight).
Quick actions in control bar
One I'm using a lot at the moment is a button placed in the 4 button persistent area at the right of the bar used to swap between input languages, I have two available: British (PC) and British (Mac). I use this because I frequently switch between using the laptop in a docked scenario, in with a regular PC keyboard and monitors and in the usual laptop-only scenario.
I'd say these two scenarios summarise the current utility of the TouchBar quite well. One, I think is genuinely cool with useful function, the other is a minor shortcut that could be quite easily achieved in other ways (like showing input sources in the macOS menu bar and swapping between them with your mouse).
Yes, I think changing volume and brightness on the touchbar is cooler, more fun and slightly more functional, but it's such a small thing that it's barely worth writing about. A big part of the problem for me at the moment is that the apps that do support it in any real way are mostly media production applications and it's not what I do. Xcode has some nice convenient shortcuts that not being an Xcode user I found kind of useful (as I don't know the keyboard shortcuts) and mail management in the built in mail app is slightly improved.
Apple are in a very fortunate position with the MacBook Pro, macOS is pretty popular with developers, and specifically the right kind of developers. The amount of attention it gets from smaller application companies is disproportionate to the marketshare it has, and that's because it has a lot of enthusiasts who use Macs and are also developers.
So we generally do see quick and good adoption of new macOS and Mac features, so I'm actually fairly confident we'll begin to see some interesting things pop up in more of the types of apps I care about (e.g. developer tools).
To conclude on the TouchBar: I'm not against it, I have some optimism for it in the future and it turns out that you quite quickly get used to not having the physical escape key, even as a developer who uses vim quite often.
Which leads me on to what for me the new hardware feature that doesn't need justifying: TouchID. This isn't a fingerprint sensor like we've had on laptops for many, many years before. Because this time it can be used by third party apps and will be used by third party apps. As a demonstration of what I was talking about above, the primary app I wanted TouchID for is 1Password. 1Password already had TouchID support ready at launch, which was fantastic news for me.
I have a long and complex master password which I mess up on constantly when I'm trying to type it out fast. Being able to unlock it easily with my fingerprint is great and I miss it a lot when I'm using the laptop in clamshell mode. I'm not so much into Apple's own ecosystem, so buying things from the App Store or iTunes with my fingerprint is less interesting, but being able to elevate to administrator privileges with macOS (in certain situations, others will absolutely require the password) with the fingerprint sensor is also a welcome convenience.
The fingerprint reader is fast, it's accurate and seems to work pretty much exactly as is expected. This is absolutely a worthwhile addition.
A source of significant concern at the moment is the battery life on these machines. So let's get this out of the way:
- The battery capacity is smaller than last generation. The devices are smaller and there is simply less room for battery.
- The improvements to chip efficiency do not make up for the difference in battery capacity.
- There's an extra display and extra systems to support, even if they are low power, it adds up. (TouchBar & the supporting systems for TouchID and the secure enclave)
When I first got my machine, I wasn't getting great battery life even during relatively light loads. 3-4 hours, which is dreadful. So I installed the then early beta of macOS 10.12.2, it helped somewhat (gaining a couple of hours in light use), but it still wasn't good enough.
The point here is that whilst there is legitimate reason for the battery life to be lesser than the older generation, there wasn't to this degree, so something had to be wrong. I decided to do a complete and total reformat and re-install of a fresh OS, no migrations of old data, a fresh start.
After everything had settled I found I was getting battery life to about what I expected. I ran a full work day and more on battery and lasted around 7.5 hours with the following workload:
- Chrome all day
- Multi-team slack all day
- Textual IRC client all day
- Microsoft Outlook all day
- Visual Studio Code all day
- Docker & 2 containers running for 3/4 of the day (one light nodejs app container and one jenkins container which built a java app every now and then)
- Twitter app all day
- Various iTerm 2 windows open all day
- 1.5 hours of spotify on and off
- Opening documents (Excel & Word) on and off
With background apps such as:
- Google Drive
It's a development load, but not a super heavy one. It's a medium load, with some periods of light load in there (e.g. when I'm mostly just reading articles on the web, chatting on slack and sending emails). The development load was primarily a nodejs app with some locally running process managers (PM2) and development of some containers, with many rebuilds (no cache) of the primary container.
I have found that kind of battery life to be fairly typical, it's what I'd call good. I can push 8+ hours with light loads (basically everything I listed above minus the development specific apps).
With heavier loads, such as constantly building and running a particular Java app in a particularly heavy IDE such as Intellij IDEA, I was getting around 5.5 hours of use - this isn't fantastic it's also not bad given the nature of the workload.
It'd probably last longer on the older generation and I think that's the main issue here. The question becomes: Is the lighter and smaller construction worth the trade off in battery?
Many would say no, I'd typically be on that side too, but if given the option I'm actually not sure if I would end up taking that side despite my many strong arguments in favour of it. The problem I have is that I carry this laptop and walk a reasonable distance with it in my bag pretty much every day, I also 80% of the time, use it docked - meaning plugged in on a desk into a couple of monitors & keyboard, so battery life doesn't really factor into my regular use - whereas a lighter laptop does.
Apple have just released macOS 10.12.2 and it has removed the battery life estimations from the menu bar battery menu, it still exists in Activity Monitor, but the reasons given are that it doesn't provide an accurate view of what a user can expect.
The rationale is that newer CPUs are becoming more aggressive and dynamic in how quickly they switch power states. This is a good thing for battery life, but using the current power consumption as a way to estimate remaining battery life becomes more difficult, as it could very much change drastically from one minute to another and provide conflicting reports that vary massively.
There is definitely, from what I've experienced in the past 6 years of using macOS, a lot of truth in that. The existing battery life estimation software is very much 'in the moment', it looks at usage currently and extrapolates, it's pretty dumb.
Why this is controversial is that the timing makes it look like Apple did this as some kind of solution to the many complaints people are having about the new MacBook Pro battery life. They almost certainly did, it may be true that this should have perhaps been de-emphasised a while back as it confuses users, but the fact it has been done now is probably for a reason.. and not a good one.
The battery life estimation is not necessarily bad if your workload is consistent over a longer period of time. For instance, I've been sat here finishing up this review and doing other little tasks, again with the app list provided above minus the development specific apps, so a light load and the estimator has been getting it pretty spot on.
It's draining in line with what it estimated earlier on (at 99%). The problem is that there are things that I could do, maybe temporarily, maybe for longer periods of time which could knock hours off that estimation or put hours on in a matter of minutes. For techies, it's easy to understand why this happens, for others, it isn't so obvious.
"One minute it said I'd get 7 hours, the next minute it said 4! What's wrong with my laptop?!"
Luckily, Apple's solution isn't to remove it entirely, just remove it from a place which would encourage obsessive monitoring of it.
As stated above, I'm running the stock upper spec model with the i7 6820HQ, 16GB RAM, AMD Radeon 455 Pro and a 512GB SSD. As far as the CPu goes, there's currently only one mobile CPU Intel makes above it, and Apple provide that as a BTO option. Performance is about the same as it has been for the past 3 or 4 years, if not a bit slower than the last haswell and maybe broadwell parts. The reason for this is likely due to the choice made by Intel to increase die space for a better (up to 20%) integrated GPU, rather than have it clock up for a low single digit percentage CPU performance increase. So they regressed slightly in terms of raw performance on the CPU side to pack in a more capable integrated GPU on a lower TDP.
Which to me seems like a reasonable trade-off, the differences in raw CPU performance are likely to be marginal and with more workloads becoming GPU accelerated, it seems smarter to invest there.
The dedicated GPU is pretty good for it's TDP (35w), I wanted something capable of playing games like Civilization VI and it does that reasonably well. Better under Windows, of course, but it does the job in macOS too. Even the Radeon Pro 455 (not the top tier model) in my machine is a significant upgrade over the 2015 model - something I believe many people are weirdly ignoring - the quite significant increase in GPU options and GPU performance.
The SSD is reaching speeds up there with the very best the consumer market has to offer (e.g. 3.1GB/s read, 2.2GB/s write), so that's all good.
Regarding RAM, depending on the price I possibly would have gone for 32GB, but it wasn't something I was looking for beforehand - was aiming for and expecting 16 and that's what I get. At the moment it's enough a good amount of headroom for what I need.
Software issues aside, it's as quick as it can be expected to be, I have other powerful PCs so perhaps it's difficult for me to be excited by it. My desktop is more powerful, with a slower (but still NVMe) SSD, so this machine was never going to surprise me in that regard.
Price and value
For me, this is the worst thing about this machine. It's stupidly overpriced and not just 'Apple tax overpriced'. It's something that made me order and cancel 4 or 5 times before finally settling on it, I knew I wanted the new MBP with the newer chipset, I had been waiting for a new MBP for well over a year, deliberately not buying the existing one.
But then it came and with a big price increase. A good portion of that price increase comes from the self-inflicted wound of the EU referendum in June of this year (£ value), but much of it also comes from generally higher prices. There are some consolations, unlike before all 15" MacBook Pros come with dedicated GPUs whereas previous generations didn't, but even given that and the touchbar (still of questionable value), it's difficult to see where the justification comes from.
Some make the argument that all they did was remove a pricing tier, which is kind of true, and the way to compare them is to compare the new baseline model with the old one with the dedicated GPU - the problem is that even at price parity the GPU in the new baseline is not as quick as the one you're comparing it to. The removal of the old baseline is really the big problem here and the additions over the old base model do not justify the price difference.
I spent a long time reviewing alternative options. My alternative options were based on the premise I'd be running Linux on them, Windows 10 is not an option, it's a bad development environment. Linux on the desktop is pretty good these days, it's quick, smooth and stable. On the desktop.
Laptops are different and whenever I came across a laptop I was interested in, the first thing I did was look up compatibility guides and forums for what does and what doesn't work when running Linux (e.g. Ubuntu) on a particular machine. The Dell XPS was a favourite, because in the US they sell 'developer editions' of both the 13" and 15".
These developer editions are essentially the same laptop with Ubuntu pre-installed. Except with some hardware differences, specifically in regards to the NIC - on the regular Windows pre-loaded versions, they come with 'Killer' NICs, on the Ubuntu/developer edition versions they come with Intel NICs. This is almost certainly done for compatibility and that alone would make me extremely hesitant to buy the Windows version to install Ubuntu on.
There are also several other documented issues with these (and many other) laptops that I don't want to deal with on my laptop. On my desktop? Fine, but then a lot of these are particular to laptops. Power management, sleep/wake, drivers for NICs, graphics switching, audio, etc, etc - no matter what laptop I looked at, there was always several wikis and forum posts about what to do to make things better and it's simply not something I wanted to deal with.
So with great frustration and anger, because I really didn't want to spend the money I did on this MBP, I returned back to what I knew would be good. I knew the trackpad would be great, I know the OS would be nice to use on a laptop (macOS still feels easier to navigate on a laptop). It didn't feel good and whilst I'm a lot happier with it these days because I know I tried to find good alternatives that I would have been happy with and just couldn't.
Due to the coupling of macOS and Apple's hardware, I feel like I should include a 'state of macOS' section in here.
It's not good. It has been up and down for the past few years and we're on a slight up at the moment, I feel, but overall the trend is down. Apple is struggling to scale and they do not pay attention to the Mac as much as they should.
We get it, the iPhone business alone is bigger than many other tech companies whilst the Mac is a small portion of Apple's revenue. But remember who supports your ecosystems, it's the people who use Macs and there's a lot of indirect revenue that is generated from those people who develop for your iOS ecosystems.
I like macOS, in many ways it's my favourite desktop OS, in others ways it has fallen behind. Windows 10 is faster and smoother on the same hardware. An issue that has persisted in macOS since the Retina Macbook Pros were introduced is low and inconsistent UI framerate, it was excusable for a short period of time, they were pushing more pixels than anyone else and downscaling on top of that to smoothen things out even more. Visually it looked great.. when everything was static.
But here we are 4 years later and poor UI framerate still plagues macOS on Macs with 'Retina' displays. It's still an issue with 2016 MacBook Pros. I repeat, this is not a hardware issue. Whilst UI scaling on Windows 10 still isn't quite as good as macOS, it's getting there and it's getting there whilst keeping the framerate high and consistent. Seriously, I boot into Windows 10 on this laptop and it's remarkable at how much smoother the UI feels. I don't want to use Windows 10, I buy Macs for macOS, but it's an undeniable advantage.
Fortunately, again, a lot my use for this machine is docked into regular monitors - where this is not an issue. I'm not running at scaled resolutions on those displays and they're driven off a significantly more powerful dedicated GPU to bruteforce past the inefficiencies and bottlenecks in macOS. But a laptop that only runs its UI well when used as a desktop replacement isn't a good laptop.
Apple was once known for paying attention to the little details, but now they're not only ignoring the little details, they're ignoring the big obvious ones too. The UI performance in macOS on Retina Macbooks whilst using it as a laptop is plain unacceptable, it's bad. Fix it.
It's way overpriced, it's not worth the money - even if you put a high value on macOS, like I do.
If you have a recent 15" MacBook Pro and you're happy with it, then you shouldn't feel the need to upgrade unless you need more GPU power or better I/O. I nearly didn't buy it because of it, but I had been budgeting for a full year in anticipation of buying a top spec MacBook Pro so was somewhat prepared, not prepared for the actual price it ended up being, but selling my old MacBook Air bridged that gap.
However, judged independently, it's a great laptop. I really like it, I enjoy using it and got what I was waiting for: Thunderbolt 3 / USB-C ports and better GPUs! Accessories need to be made available, either via USB-C or Thunderbolt 3, but they're coming (many are already available), there are over 50 laptops with Thunderbolt 3 ports and that's growing all the time.
For the 100th time, as someone who does use this in a desktop scenario quite a lot, the Thunderbolt 3 and/or USB-C expansion capabilities are primarily what made me wait. It's here, I'm happy and I'm excited for some thunderbolt 3 docks, multi-display adapters, etc. It lengthens the life of a PC in this use case, I don't need to worry about being able to run multiple high resolution displays in-tandem with other high-throughput devices in the future, and given the wider adoption of Thunderbolt 3 already, not having to worry about using adapters for Thunderbolt 2 devices (which differ more in more ways than just speed).
I love the design, it looks and feels nice and I'm genuinely a fan of the keyboard and trackpad. All in all, it's a Macbook Pro, I mostly got what I was expecting.