NotebookCHECK - Notebook Forum

I think the reported RAM speeds are wrong for 13" without touch bar "LPDDR3-1866":
https://www.apple.com/uk/macbook-pro/specs/

You are correct, I fixed the specs. thank you!

While Apple may be boasting about their impressive sequential read and write speeds of the SSD, their claims don't seem to match up. According to the small text at the bottom of the product page they achieved up to 3.2 GB/s read and 2.2 GB/s write sequential speed with FIO-2.19 (best case scenario). Knowing that a bigger SSD will help, still the 256 GB that is tested here is quite far of the mark.

Now the CrystalDiskMark and AS-SSD benchmarks have been performed on Windows probably via bootcamp. It could be that due to 'not so optimized' drivers specific to the storage controller could result in less extracted performancein Windows. It would be interesting to find a similar storage benchmark suite that would run natively in MacOS and compare the results.

Many other reviewers have all boasted the same PR talk telling how amazingly fast the SSD is without actually some serious benchmarking. Yes the sequential speeds are still very respectable. But the 4k Read and Write metric which is also very important and probably very representative of regular use is awful compared to the competition. Apple really screwed that up.

Also provided all things equal, the CPU should be slightly more efficient, but less battery runtime compared to the previous model. That is a huge bummer. What is Apple doing?

Keep up the good work! I will be following along the updates.

Can you please review the base model Macbook Pro 2017 (2.3Ghz i5) ?

Hi, I'm a master in British and my major is computer science, I wanna to buy a macbook pro 15 now and my expectation is 2.9Ghz i7 and 512G SSD, I don't know whether the configure is enough or not because I want to use it longer for three or four years. So could you give me some advice? By the way, do you know the runtime of this configuration? Because sometimes I need to go to the classroom and prepare my paper, so it is necessary for me to have a longer battery runtime.

Hi, may you please check the new keyboard when machine is warm/hot?
The problem like "Keyboard Clicking/Sticking When Machine is Warm" happened on the mbp 2016 late.

Very good review as expected. One minor correction required though:

"Apple advertises a performance boost of 50% in the new MacBook Pro 13."

This is only the case for the "MacBook", therefore it means only for the 12" model, the SSDs are upgraded.

Hi. I'm also very interested in the non touch bar (function keys) 13" macbook pro. Is it too much asking a deep review on the 2.3 GHz unit like the one you did with the 2016 2.0GHz MacBook Pro?

I'm very interested in the Battery Life aspect of the 2.3GHz, because I doubt the 7360U Kaby Lake CPU -with a max frequency of 3.6GHz- and the new 2133MHz Memory RAM, will keep the same autonomy with the exactly same 54.5W/hour battery.

Also, I'm interested in the heat of this unit, again because of its much higher clockspeed both minimum and Turbo boosted.

I'm very interested, because right now there are discounted 2016 nTB MacBooks form 2016, and due to its incredible battery life and low speed (and low temperatures) are of interest for many of us.

Also, it would be nice you to check the keyboard and compare both, the 2016 and 2017 just in case Apple made some small tweak to it.

But mainly, battery life on web browsing and battery life on full speed, and check out any throttle issue, please.

Thank you in advance.

The other MacBooks of the 2017 lineup are also planned for review, but we might wait for our partner shops to get them in stock for us to review. Sadly no review samples from Apple directly currently, but on the other hand that means no dedicated press samples that might be cherry picked, so stay tuned.

I'm conflicted. How does the touch bar model with the more efficient kaby lake
processors have worse battery life? That doesn't make any sense. This is
frustrating because I expect excellent battery life from the 13 inch laptops - I'd gladly purchase the non touch bar model if not for the 4 ports, touchID, faster wifi etc. (not sure whether the touch bar is useful or not). By the way, what brightness level does 150 nits correspond to? The macbook pro has a maximum brightness level of about 550 nits - shouldn't the test be performed at 75% brightness like apple does their tests?

150 nits is -4 brightness settings (of 15), so quite similar to the 75% and we of course use a comparable brightness level across all devices. Otherwise dimmer laptops would have an advantage.

Apple is selling refurb 2016 2.9 Ghz MacBook Pros for $400 less than a new 2017 model. Any comments on whether the improved performance is worth $400?

Quote from: Alfred on June 19, 2017, 00:52:11
I'm conflicted. How does the touch bar model with the more efficient kaby lake
processors have worse battery life? That doesn't make any sense. This is
frustrating because I expect excellent battery life from the 13 inch laptops - I'd gladly purchase the non touch bar model if not for the 4 ports, touchID, faster wifi etc. (not sure whether the touch bar is useful or not). By the way, what brightness level does 150 nits correspond to? The macbook pro has a maximum brightness level of about 550 nits - shouldn't the test be performed at 75% brightness like apple does their tests?

The touchbar models uses 28 Watt processors and uses 2 fans to cool the chips whereas the non-touchbar models uses 15 Watt processors and 1 fan to cool the chips.  You can see how the touchbar model may feature less battery life given it is nearly double the TDP of the touchbar model and may or may not feature a smaller battery.

(part in response to Apple SSD comment and also points out why Apple may prefer unusual CPU choices with wider clock operating range)

This site does great job already but some of the commenters may wish to start thinking things more broadly:

eg. How long does say 3 LBS device stay usable, using wifi at outdoor readable brightness, viewing say Twitch stream. Or what about streaming into twitch?

Why I say that? Well if you consider that complete scenario, it may well turn out that things like CPU speed and SSD speed may need to actually be "regulated" such that benchmarks (unless detected and then regulation turned off... then you get "cheating!") may turn out poorer results.

eg. in this sort of scenario, if CPU clock speed is not capped, it can end up running at higher clock than what eg. twitch use actually needs. Lots of reviews indicate that 7100u that has "clock cap" and no turbo but does clock down of course, does fare better in these sort of "tricky balance scenarios" than some higher end models that end up turbo boosting when not really necessary. To further validate this I ran twitch several hours at default settings and then manually capping the clock at 1.2 Ghz instead of 2.4 max. The stream performed the same but battery life did improve measurably. (However I think bios update may have fixed something that was causing it to clock up more easily but this illustrates the point anyway - if a blend of background tasks or tabs open would have javascript in background, clock could stay above 1.2 Ghz while the foreground window task didn't actually need more than 1.2 Ghz).

So CPU,SSD, Wifi etc speed may be useful to regulate even if benchmarks suffer - most (=none) PC reviews don't normalize these benchmarks to things that you'd have to for proper balanced scenario review - such as color quality at high yet standardized brightness, white uniformity, audio quality from speakers, network latency, battery life.

ie. How much battery life you get while maintaining acceptable network latency, foreground tasks running smoothly, background tasks not causing cpu or ssd to "hurry up" for them since who cares, they are in the background. Oh wait. PC tech reviewer is the only guy who cares since they run 10 benchmarks simultaneously in a hurry to get review out. Wake up I say. That stuff only applies on desktop use.

Now to make things REALLY complicated, which device performs well in mobile studio use? Now you have to maintain low latency I-O with battery life for pen, musical keyboard input etc. Same for racing games (TMNF with 10 ms input lag is too much).

Not complicated enough? Well how about vacuum testing? I return devices if they smell like China waste dump. Not all the chemicals smell but can still affect people if not in ventilated office (residential standards don't have enough ventilation in most countries to cover for a lot of synthetic outgassing materials indoors - retail stores have massive amount of ventilation to cover up the china smell - or they would if the china stuff wasn't wrapped in air tight plastic which transfers the problem to end user).

I would add to previous post that if there is some "regulation" that knocks say SSD well below competition like in this Apple SSD case, it sure would be nice to have a way to say "I'm benchmarking or doing intense performance task on battery" .. In Windows you can set Performance mode, not sure about Mac. And not sure if it affects SSD perf in Windows either (should regulating that be of help).

This is so that some people may want to bench the "classic" singular scores, for score purposes but "complex people" care about complex scenarios like, is the mobile device actually usable in mobile use and for how long and that involves different metrics and power use profiles etc.

Hi! Do you know the AdobeRGB % coverage?? Thks!

QuoteThe other MacBooks of the 2017 lineup are also planned for review, but we might wait for our partner shops to get them in stock for us to review. Sadly no review samples from Apple directly currently, but on the other hand that means no dedicated press samples that might be cherry picked, so stay tuned.

Thank you Klaus, provided that the new Kaby Lake CUPs seem a bit more power hungry, this will be crucial for the 2.3GHz model battery life. Also, I'm a bit startled with the 2133MHz DDRL3. Those are'n supposed to be DDR4 frequencies? Will this rise in the RAM speed affect the power consumption? Remember the 2016 without touch bar carries 1866MHz RAM.

All this make the new Kaby Lake update a bit of a mystery.
Eager to read your 2017 2.3GHz MacBook Pro Without Touch Bar review.

The recent update completely removed colour accuracy data in sRGB mode (which is far less accurate than P3 mode by the way). Why??

Also it's bizarre that 2017 model uses PWM instead of DC as seen in all previous models.

As some others stated in the comments already, I am also interested in whether the keyboards on 2017 models are improved over the late 2016 model, i.e. do they still have those issues when the machine is warm?

QuoteOur MacBook Pro 15 (Late 2016) we use in the office also had some problem with the "S" key recently. The key stroke was either not triggered or triggered twice. However, the problem was gonge after a while.

This sounds very likely to be the said issue, quite common on 2016 models, and it only happens when the machine is heated up.

Also, gonge -> gone.

There is one person on reddit who had her/his 2016 MBP replaced four times with the same keyboard issue, and when she/he finally got the 2017 version, the issue hasn't appeared so far.

Could you also test this in your review please?

I'm also quite puzzled about the fact that this model has worse battery life than the previous generation!

I waited for the 2017 model specifically to ensure that the battery life issues with the previous models were overcome since this packs a more efficient chip.

Could this problem be related to the specific review unit?

Many thanks

We got another model with 512 GB in house and it also showed comparable battery runtimes. So either the old MBP13 battery runtime was somehow too long (e.g. adaptive brightness, altough I dont think so) or the updates in OS X also did change something. Kaby Lake itself should be more power efficient.

Regarding the s-key problem, that happened with our 15-inch model after months of usage, so its hard to judge after 2 weeks if the new keyboard is better in that regard.

I intend to buy 2017 (KabyLake) 13inch with touch bar, mainly for music production and live playing (Cubase, Ableton Live, Kontakt, Plugins etc.).

I will definitely get 16GB ram and 512GB SSD, but wondering whether to go for the i5 or the i7 configuration?

Be glad to get your smart advice.

Tnx

Hi! Will be much more delayed your non-touch bar 13" MacBook Pro review?

I'm in a bit of a hurry, because the units of the 2016 are running out of stock everywhere, and I need to make the decision ASAP.

Basically I need the best battery life, and as I read, the new Kaby Lake processors may be a bit more power hungry. This machine for instance, performs worse battery-wise. And that's only explainable due to its higher clocked Baby Lake CPU.

So I need to reed your battery test and comparison of the 2016 ntbMBP and 2017 ntbMBP in order to decide wich one to buy. If the battery performance is the same -or better- then I'll pick up the newest model. If it is worse, which is probable, then I have to hurry and pick up one of the 2016 models.

Just give us a deadline, if it's not too much asking, in order to wait or not for your review (or, at least, the preview). The next week, maybe?

Thank you in advance.

Thanks for the good detailed review.  I'm also trying to decide between this touch bar vs the non touch bar version.  Is that initial review going to be posted soon?

Non Touch-Bar is in house, first informations will be out next week. Regarding the keyboard problem, it definitely was not related to a heat up unit, but was happening all the time.
Regarding i5 vs i7, it usually does not pay off to take the higher end CPU, other than if you really need every MHz of speed (e.g. rendering tasks) or you dont care on money.

The 2.4 GHz problems appears wičh other laptops too, with particular USB A 3.0 card reader, i think it is caused by bad shielding of that reader cable. For example, other USB-A 3.0 accessories are fine (does not matter whether it uses usb cable or directly connects to port like thumbdrive),bluetooth stops too,  strongly depends on the distance to the display

Are you sure this is problem of the notebook itself?

Can you please test the color accuracy for sRGB on this and the non touch bar version?

Well testing sRGB accuracy is a bit wonky in OS X with Calman Client (that we use). I am not sure if the results are really accurate as the tools are switching profiles when testing. Thats why we sticked to P3.

waiting for macbook pro 13 inch 2017 non touch bar review...

Do you expect the non touch model to have the same colour accuracy? As the 2016 model appeared different between touch and non touch. Do you think that might have been from the reading error?

Will P3 colour accuracy be indicative of sRGB colour accuracy?

Will mbp 13 sRGB be more colour accurate than on the dell xps 13?

Unfortunately, these scores are unreliable and underscored until tested with CrystalDiskMark NATIVE for Mac, for obvious reasons. Thus, the article is utterly misleading.

Your reviews are incredibly good, Andrew. Thank you so much. Really, they are unparalleled and so helpful.

Maybe you could retest the PCMark 8 Storage Score v2 with macOS 10.13/10.14/10.15?

It seems like the performance there has taken a dive in later MacBooks, and I wonder if it is just APFS, or something weird in Apple's newer hardware.