I linked to my customer review that compares the 2016 and 2015 15" models, but someone thinks that's against the rules, so here's the material without the link:
I did a bit of research before picking a computer, and I'll put some of it here in the hope it may help others. I'll start with my somewhat subjective experience, and then I'll do a comparison based mostly on more objective things. I think there are better values for many people, but the new MBP is a good deal for some. It all depends on what you think is worth paying for, of course.
I'll compare to the two alternatives I think are the most direct competition to the new 15" MBP: last year's 15" MBP and the Dell XPS 15 9550 UHD.
My experience with the new MBP so far (updated after a month)
I got the 15" 2.7 GHz 512 GB 455 dGPU model, replacing a 2011 MBP with a matte 17” screen. It will be my only computer for daily use online and editing photos. I got the more expensive of the two 15" configurations Amazon sells mainly for the larger, faster hard drive, though the faster processors will probably be handy sometimes too.
My overall experience of the screen is very positive. Bright, sharp, colorful, noticeably more of each than before. Naturally, I miss the extra screen space, but the sharpness of this display does help make up for the loss.
Word is that Apple’s glossy screen is less reflective than most, but it’s still a lot more reflective than I’d like. It hasn’t been a serious problem yet, using it indoors, but I can see that in some circumstances I’ll have to turn the screen brightness up more than before to overpower reflections, and some I'll just be stuck with.
The display looks evenly illuminated when I put up a full white screen.
Everything is running smoothly, on the whole. I have 16 browser tabs open, listening to a radio station on iTunes, watching an HD TV stream, writing this in Text Edit, and everything is zippy. Can't hear the fans, the computer isn't even warm. I'm also doing some stuff in Affinity Photo. Nothing too intense yet, but normal editing is fast.
I might occasionally do all but the photo editing on battery power, so I ran a test of my typical use with those applications except Affinity. I got over 11 hours before I plugged back in at 5% power remaining, including four hours of basketball HD streaming video. The screen brightness was low, the way I usually have it indoors. I turned it up a little for the basketball.
It took just under an hour to recharge to 75%, 1:45 to recharge completely.
Out of the box I ran the machine for a week without restarting before it started getting balky and I restarted it. It's been very stable, with only a couple spinning beachball moments that have cleared themselves up in a minute. I've restarted one more time in the month I've had it, with one spinning beach ball that was resolved by closing a tab and reopening it. Excellent for a new model.
Touch ID (fingerprint recognition) has been convenient, faster and easier than passwords. No hitches.
The touch bar is nothing special for me, at least not yet. I haven't really tried to learn the helpful things it can do. I do like skipping Youtube ads. I may try out the piano app. It’s a new thing that nice stuff will be invented for. My only gripe is that it goes dark after a minute of touch bar/keyboard/touchpad inactivity, so then you have to touch something to wake it before using it. (That may be to avoid screen burn, a problem for OLEDs.)
I really like the keyboard. It's quiet, or a lot quieter than I expected after reading all the complaints online. (Hah, the "ClackBook Pro," someone said.) If you pound on it, it’s loud; if you don’t, it isn’t. The shallow travel of the keys is fine with me. The backlight is even, and doesn’t leak around the keys much, much less than for the 2011.
The huge trackpad works well, but I'm still getting used to it. The size allows finer control of the cursor and larger gestures. I haven’t had trouble with accidental clicks since I disabled touch-to-click. it does add an extra click once in a while, doesn't affect how it works but is distracting. I’m having no problem with the three-finger gestures, unlike some. The only problem is remembering which gesture does what. Like the keyboard, clicking isn't as loud as some have said, but the quiet click sound is sometimes a bit tinny.
The speakers are roughly like having a quality pair of portable bluetooth speakers built in, which is to say, they’re impressively crisp, clean, bass shy, and loud. In my quiet house I can go a couple rooms and 50 feet away and still listen to the news. But I’m glad to have the headphone jack for serious music or movie listening.
I never used many ports, so I'm not broken up that they're gone. I’m using
a small cheap adapter with good results for the old USB stuff: DVD player, wireless headphones, hard drive, thumb drive. Works with everything I've tried it with, and it's small and cheap enough that I just leave one one the end of each cable I need one for.
I love the concept of the MagSafe charger cable, but with my 2011 I eventually had trouble with it showing it was connected (green light) when it was actually wasn’t, which resulted in unexpected battery drains and shutdowns. The plastic coating over the wire also decomposed, and after much tape I finally had to replace it (about $80 new). I found there are actually a lot of complaints about MagSafe when I shopped for the replacement. So I don’t miss it as much as many people do.
I got the silver color not because it looks better now, though it does look very good, but because reports suggest it will look better than the space gray once it gets scratched up.
The build quality appears excellent. It feels sleek, precise and solid.
**********************
Here's how the new MBP compares to its nearest competitors, as I see it. (I'll elaborate on each point in the first comments to this review.)
Comparison to 2015 MBP 15"
The new MBP 15" is clearly better than the 2015 model in:
-- Screen
-- dGPU
-- SSD speed
-- External monitor support
-- Heat
-- Quietude
-- Speakers
-- Touch ID
-- Touch bar
-- Size/weight
It's also mostly better in regard to:
-- Battery life (see third comment to this review)
Mixed:
-- Ports (the new machine has by far the more powerful and flexible ports, but the 2015 has by far the more convenient ones for most people not yet living in the future)
-- Keyboard and trackpad (the new ones get mixed reviews compared to the old)
The one area in which the 2015 model has a clear, unmixed advantage is:
-- Replaceable/upgradable SSD (and if the motherboard fails, you won't need Apple's help to recover your data)
*
Comparison to Dell XPS 15 UHD
The areas in which the new MBP has a clear advantage over the XPS UHD are:
-- SSD speed
-- Battery life
-- External monitor support
-- Noise (probably)
-- Speakers
-- Size/weight
Many find the MBP to have a better build quality, but not all agree.
Mixed:
-- Screen (UHD display beats the new MBP's in resolution, but it lags in brightness, contrast, and initial color accuracy--though the last point can be fixed with additional equipment)
-- Processors (CPU and GPU are on a similar level)
-- Ports (as with the older MBP, the Dell has the much more convenient selection for most people, but they're much less powerful and flexible)
-- Keyboard and trackpad (comparisons remain controversial)
A major consideration for many is the operating system, but some will prefer one and some the other. That the Mac can run Windows fairly efficiently is a plus.
Many would consider the Dell's touch screen a clear advantage, but some few will prefer the Mac's touch bar. There's no doubt the touch screen can do much more useful things, but I'd be reluctant to use it myself. I don't like marking up my screen, and even a teflon pen does that over time. The touch bar is better that way.
Some clear advantages for the XPS:
-- Allows up to 32 GB RAM (well, at the cost of substantial battery life loss, even for the 16 GB version, which uses the same desktop RAM--see third comment to this review)
-- Repairability and upgradability
*
A 2015 MBP 15 with features comparable to the new MBP base model with a 512 GB SSD was $2500 when it came out, $100 less than that configuration now. The comparable XPS UHD was $2100, $500 less. Now you can gets deals on refurbished 2015 MBPs with dGPU (be aware that many being sold now don't have a dedicated GPU), and the price of the similarly equipped XPS is down a couple hundred dollars or more since it came out.
So is the new MBP overpriced? It depends on what you value. Its price will come down as those of its predecessors did, but even at the opening price it's a strong offering (as sales suggest). It's clearly better than the 2015 model in most ways, ways worth paying for for many of us. I think the Dell is an outstanding deal, and I can see why many would prefer it, in part for its lower price. But even there, I think a thorough comparison shows why for some the new MBP will be the better deal.
***********************
The numbers I give below are based on actual measurements drawn from professional reviews or professionally curated aggregates of user measurements. I'll give more particulars and links in another comment to this review, as I have time to do it.
Screen
The new 15" MBP screen is one of the best available for a laptop. The resolution is the same as last year's, 2880 x 1800 pixels, but the brightness and contrast have improved almost 50%, the range of color has increased over 25%, and the accuracy of the color has also improved, to the point that differences from the target colors are mostly unable to be perceived by humans. Last year's screen was already excellent, but this one is significantly better.
The comparison to the Dell XPS 15 UHD is more complicated. It too has one of the best screens available. It's 4K, 3840 x 2160 pixels, which is useful if you want to edit or view 4K video at full resolution. However, the new MPB screen has 20-50% more brightness and contrast (depending on who measures it). The color accuracy of the Dell is worse out of the box, but it can be calibrated to excellent levels, if you have the equipment needed.
A more subtle difference is that the Mac aims at the P3 color range, used in digital cinema and video, while the Dell aims at the Adobe RGB standard, used in still photography. Normally you wouldn't be able to see any difference, but at the extremes of color saturation the Mac should be able to show more reds, magentas, and red-oranges, while the Dell should be able to show more greens and cyans.
CPU
The main processor of the new 15" MBP is actually a little slower for some tasks than last year's model, but faster for others. It's more or less a wash for most ordinary stuff.
The Dell XPS 15 UHD has the same processor as the new base 15" MBP, so it's on a similar level, but the new MBP offers optional higher base clock speeds, 2.7 or 2.9, that can give it an edge. By a standard comparative benchmark (Geekbench 4--details will be in next comment), the 2.7 is 8-11% faster than the 2.6, and the 2.9 is 7-9% faster than the 2.7. (Amazon only sells the 2.6 and 2.7 directly, though Marketplace sellers may offer the 2.9. You can all sorts of configurations directly from Apple.)
There have been a couple direct comparisons of the new and old MBPs that have found that for some heavier jobs, the new machine is significantly faster, up to 90% faster for some tasks. This may be due almost entirely to the better dGPU and faster SSD, and/or related "throttling," all of which I'll get to below. Links:
Jonathan Morrison compares the new 15" MBP and its 2015 counterpart. The part I referred to starts at about 2:30:
Apple Insider's comparison with similar results:
http://appleinsider.com/articles/16...acbook-pro-with-touch-bar-vs-2015-macbook-pro
See also the version specifically about video editing, starting at 3:50:
Excessive CPU throttling was a problem for the XPS 15 UHD when it came out, but firmware updates have fixed that.
GPU
The new MBP 15" has a choice of three dedicated graphics processors, the AMD Radeon Pro 450, 455, or 460. By a standard benchmark test (Geekbench 4--details will be in next comment), the 455 is about 15% faster than the 450, and the 460 is about 10% faster than the 455. (Amazon offers configurations with only the first two: the 450 comes with the 256 GB model, the 455 comes with the 512 GB model. Marketplace sellers may offer other configurations including the 460.)
The 2015 MBP offered the Radeon M370X dGPU. The new 450 is about 9% faster than the M370X, and of course the 455 and 460 are even faster. That's undoubtedly one reason the new MBP is faster at heavy-duty graphics tasks.
Another reason is heat. Apple discontinued the sale of new 2015 units with the M370X, in favor of units with only a much inferior integrated GPU, apparently because of problems caused by overheating. Excessive heat causes "throttling," which is an automatic slowing of the processor when some parameter like temperature is exceeded. If it gets too hot it can also cause hardware failure. The GPU chips in the new MBP don't run as hot, so there's less throttling.
The XPS UHD has an NVIDIA GeForce GTX 960M dGPU. It's about the same speed as the 455, so between the 450 and 460. It's throttled for power conservancy when operating on battery, so graphic-intensive work on the go will be limited.
Gaming****************
I'll get to this later.
RAM (see third comment for details)
One of the biggest complaints about the new MBP has been that it's limited to 16 GB RAM, as the 2015 model also is. The Dell, on the other hand, can be expanded up to 32 GB. Most people don't need that much, but if you do need it, not having it can really slow things down. And it's great knowing you have it just in case.
But there's a trade-off. More RAM uses more power, and since laptop processors don't accept 32 GB low-powered RAM yet, it also requires more power-hungry desktop RAM. The RAM Dell uses draws over 40% more power than the RAM the Macs use. How much that matters depends on how much you use your RAM (for, say, video editing or Chrome).
I'll do the math in the third comment to this review, but as a very rough guide, in ten hours it would cost at least 5 and potentially as much as 84 watt-hours extra power for 32 GB. Considering the current battery is 76 watt-hours, and the largest you're allowed to have on a commercial flight is under 100 (FAA rule), it can add up fast.
Using the desktop RAM would also make the 16 GB version less efficient, costing 2 to 18 watt-hours in ten hours.
Storage
The SSD storage on the new Mac is reportedly the fastest available by a large margin. This isn't just a matter of using the fastest Samsung parts, but also Apple using their own custom controllers. The new SSD is very roughly 30-50% faster than the old MBP one, and about twice as fast as what comes on the XPS.
The disadvantage of the new MBP 15" SSDs is that you can't upgrade them. They're soldered onto the motherboard. You can upgrade with the old MBP and the XPS.
The fact that the SSD is stuck on the motherboard means if the motherboard fails, probably only Apple will be able to recover the data from it, using a special connector. Makes backups even more important.
Battery life (see third comment for details)
There have been complaints about the battery life of the new MBP. Some who complained reversed course after an OS update, though Apple claims it did nothing to change battery life. Professional reviews have generally found the new 15" matches more or less Apple's 10-hour claim for light use such as web surfing or video playback--some found the machine exceeded it.
In the tests I've found where precision and uniformity of testing conditions were attempted, the new MBP has consistently done as well as or, usually, better than the 2015 model and the XPS UHD in light to medium use. The advantage of the new MBP has ranged from a little to over twice as much run time, despite having a smaller battery. I have seen some impressionistic reports to the contrary, but I put more stock in more precise and controlled tests. In some heavy-duty use, the larger batteries of the 2015 MBP and the XPS UHS do better. (I don't count early tests of the XPS, because its battery life was later improved by BIOS updates.)
Ports
The new MBP has two kinds of ports: an earphone jack and 4 Thunderbolt 3/USB-C ports. This is known to some as simplifying/maximizing the power per port, and to others as Donglegate. Each Thunderbolt 3 port can do more than several more old-fashioned ports put together, but you will need adapters, a hub, or new cables for most existing devices, even Apple devices.
The older MBP has 2 Thunderbolt 2/Mini DisplayPort ports, 2 USB 3.1 gen 1 (formerly known as USB 3.0), 1 HDMI, an SD card reader, a headphone jack with optical output, and a MagSafe 2 charging port.
The XPS UHD has 1 Thunderbolt 3 port, 1 USB 3.1 gen 1 with PowerShare, 1 HDMI, an SD card reader, a headphone jack, and a power receptacle.
The main differences between Thunderbolt 3 and 2 are that the 3 has twice the capacity (40 GB/s), its USB 3.1 gen 2 has twice the capacity of USB 3.1 gen 1 (10 GB/s), and it can power the MBP. Both allow up to six devices to be "daisy-chained," i.e. connected serially.
External monitor support
The new 15" MBP can drive two 5K or four 4K monitors at 60 Hz at the same time (!).
The 2015 15" MBP with the dedicated GPU can drive one 4K or 5K display at 60 Hz, or two 4Ks at 30, or three displays of sundry sorts. Plus one more of moderate resolution via Apple TV, I've read.
The Dell can drive a 5K or two 4K monitors at 60 Hz, but only via an optional docking station. It can presumably drive multiple lower-resolution monitors with its various ports.
Heat
The new MBP runs cooler than last year's model, which has some trouble with thermal throttling under heavy loads. The Dell also had heat and throttling trouble when it came out, but that was fixed by a firmware update.
Silence
Quiet fans have long been a thing for Apple. The noise levels of the new MBP, the 2015, and the XPS are similar in terms of decibels. Apple has taken new measures to make the sound less annoying, though, using redesigned split blades on two fans that run at slightly different frequencies. That's in addition to the uneven spacing of blades they've been using since 2012. It all helps prevent noise peaking at one pitch, so it sounds more like white noise and is less noticeable. It seems to work, judging from the frequency response graphs and subjective reviews.
The new model also sounds better than the previous one because it doesn't heat up as easily, so the fans don't spin up as much.
Speakers
The new MBP speakers can get loud, measured at 86 dB. Actually measuring speaker performance for laptops seems to be a new thing, so I can't compare by the numbers, but I have seen video comparisons of the new MBP speakers and the 2015 speakers. Not a close call, the new ones are louder, less tinny, fuller. See for example starting at 1:45 in this video that was mentioned above:
http://appleinsider.com/articles/16...acbook-pro-with-touch-bar-vs-2015-macbook-pro
I haven't seen any direct comparison to the Dell, but I've noticed people are more effusive in their praise of the new MBP speakers than of any others in a similarly-sized machine. People do like the XPS's speakers, but some have complained that their downward-facing design is easily muffled.
Both the new MBP and the Dell had speaker crackling problems, to the point of failure with some MBPs, but both were fixed with software updates.
Touchy subject
Only the Dell has a touch screen. Only the new MBP has a touch bar. The touch screen is a lot more useful, but using it leaves fingerprints and smudges on an otherwise beautiful display. The touch bar gets that too, but it doesn't matter so much (and oddly doesn't show in most conditions).
Keyboard
The new MBP has a low-travel keyboard, i.e. the keys don't press down very far, only .5 mm, which has caused great consternation and reportedly injuries to some (seriously--claims of bleeding fingers and, judging by comments, a few heart attacks). But it's greatly loved by others. The old MBP and the XPS have about 1.5 mm travel. If you think that will matter to you, you'd better try it out before purchase.
Trackpad
The new MBP trackpad is huge, a lot larger than those of the old MBP and XPS. That allows finer and faster control of the cursor, and better use of multi-finger gestures, but it can also lead to more unwanted inputs by palms or whatever. The benefit-to-trouble ratio varies a lot from person to person, but most are liking it, based on reviews and comments.
Repairability/upgradability
Hah hah hah, repairability-schmepairability, says Apple! The new MBP may be the least repairable laptop out there, with the RAM and SSD soldered to the motherboard, the battery glued to the case, and the touch bar glued in too. Apple also uses unusual pentalobe screws to make access difficult. On the bright side, if the motherboard fails, Apple can recover your data from the SSD with a special tool, and no doubt a special fee.
The 2015 MBP 15 wasn't a champion of repairability either, but you can replace or upgrade the SSD with a pentalobe screwdriver.
The XPS 15, on the other hand, is highly repairable and upgradable. In particular, the RAM, SSD, and battery can all be replaced at home with fairly ordinary tools.
Size/weight
The new MBP is smaller and lighter than the 2015 MBP and the XPS 15, though the XPS is a bit less wide.
New MBP: .63 x 13.74 x 9.49 in (16 x 349 x 241 mm)
3.98 lbs (1.807 kg)
power supply: .81 lbs (369 g) = 4.79 lbs together
Old MBP: .71 x 14.13 x 9.73 in (18 x 359 x 247 mm)
4.49 lbs (2.04 kg)
power supply: .71 lbs (320 g) = 5.2 lbs together
XPS: .67 x 14.06 x 9.25 in (17 x 357 x 235 mm)
4.58 lbs (2.077 kg)
power supply: 1.01 lbs (458 g) = 5.59 lbs together
OS
I'm not going to go over the advantages and disadvantages of Mac OS vs Windows. But it's worth noting that you can run Windows on a Mac a lot more effectively than vice versa. It's common for Mac users to run Windows to play games and run other programs that do better in Windows.
Style
I think we all know who wins here.
Just kidding. Not going to say anything more about this except that all three machines are handsome and one should consult their own taste.
**************************
Some helpful reviews:
2016 15" MBP
http://www.notebookcheck.net/Apple-MacBook-Pro-15-Late-2016-2-6-GHz-i7-Notebook-Review.185254.0.html
http://arstechnica.com/apple/2016/11/the-2016-13-and-15-inch-touch-bar-macbook-pros-reviewed/
https://www.engadget.com/2016/11/14/macbook-pro-review-2016/
http://www.pcworld.com/article/3143635/computers/surface-book-i7-vs-macbook-pro-fight.html
http://www.mobiletechreview.com/notebooks/2016-Apple-15-inch-MacBook-Pro-Touch-Bar.htm
http://www.expertreviews.co.uk/apple/apple-macbook-pro/page/0/1
http://appleinsider.com/articles/16/11/14/review-apples-late-2016-15-macbook-pro-with-touch-bar
http://thesweetsetup.com/articles/review-15-inch-macbook-pro-touch-bar/
2015 15" MBP
http://www.notebookcheck.net/Apple-MacBook-Pro-Retina-15-Mid-2015-Review.144402.0.html
http://arstechnica.com/apple/2015/0...e-2015-15-inch-retina-macbook-pro-reviewed/2/
http://www.mobiletechreview.com/notebooks/15-inch-Retina-MacBook-Pro-2015.htm
http://www.laptopmag.com/reviews/laptops/apple-macbook-pro-retina-15-inch-2015
Dell XPS 15 9550 UHD
http://www.anandtech.com/show/10116/the-dell-xps-15-9550-review
http://www.notebookcheck.net/Dell-X...HD-InfinityEdge-Notebook-Review.156462.0.html
http://arstechnica.com/gadgets/2016/01/dell-xps-15-review-a-bigger-version-of-the-best-pc-laptop/
http://laptopmedia.com/review/dell-xps-15-skylake-9550-review-every-designers-wet-dream/
https://www.engadget.com/2016/04/06/dell-xps-15-review/
http://www.mobiletechreview.com/not...e-2015-skylake-the-ultimate-windows-10-laptop
http://www.windowscentral.com/dell-xps-15-9550-review
http://4k.com/laptop/a-review-of-th...ltra-hd-touch-screen-laptop-7000-7548-series/
http://www.laptopmag.com/reviews/laptops/dell-xps-15
http://www.ultrabookreview.com/10234-dell-xps-15-9550-review/
http://laptops.reviewed.com/content/dell-xps-15-9550-laptop-review
Details
Actual performance often varies from specs, and machines and measurements vary from each other, so I've gathered some numbers for comparison.
Screen:
Brightness (average across screen, in nits)
-- New MBP 15"
---- 440 (Laptopmag)
---- 451 (Mobiletechreview)
---- 465 (Notebookcheck)
-- 2015 MBP 15"
---- 300 (Mobiletechreview)
---- 303 (Laptopmag)
---- 315 (Notebookcheck)
-- XPS 15 UHD
---- 285 (Laptopmag, an outlier here)
---- 334 (Anandtech)
---- 344 (Notebookcheck)
---- 345 (Laptopmedia)
---- 350 (Mobiletechreview)
---- 363 (Expertreviews)
LED displays typically have some unevenness of brightness across the screen. This can vary a lot from one sample to the next. Notebookcheck gives the measurements for different parts of the screen, which they divide into a 3 x 3 grid. The sample of the new MPB 15" they tested had differences of just under 10% between different parts of the screen, from 444 to 483. That's quite a bit better than their MBP 2015 sample, which had a range of over 20%, from 279 to 341. The XPS was measured from 325 to 361, a difference of 11%. Laptopmedia also tested the XPS and found a range of 326 to 367, 12.6%.
Black levels (nits)
-- New MBP 15"
---- .33 (Notebookcheck)
---- .37 (Mobiletechreview)
-- 2015 MBP 15"
---- .34 (Notebookcheck)
-- XPS 15 UHD
---- .32 (Anandtech)
---- .34 (Expertreviews)
---- .38 (Notebookcheck)
Contrast ratio is the ratio between the brightness of darkest black--the black level--and brightest white, both measured with the screen set at a single brightness.
-- New MBP 15"
---- 1180 (Mobiletechreview)
---- 1464 (Notebookcheck)
-- 2015 MBP 15"
---- 1003 (Notebookcheck)
-- XPS 15 UHD
---- 950 (Notebookcheck)
---- 980 (1140 before calibration, Laptopmedia)
---- 1022 (Anandtech)
---- 1065 (Expertreviews)
Color range (space, gamut)
Measuring and talking about color range is complicated by differing standards, some of which are misleading. Often which standards are used in a claim about color range aren't specified. I'll give a few figures that I think stand the best chance of being reliable.
Notebookcheck measures the new MBP 15" as showing 99.1% of its target P3 color range. P3 includes all of sRGB, the old standard, plus 26% more. The 2015 MBP 15" shows 91% of sRGB, according to Notebookcheck (so in theory the new model should show at least 37% more than the old).
Oddly, Notebookcheck measures the XPS UHD as showing only 89% of the Adobe RGB range, while the other reviews I've seen put it from 97 to 100%. Laptopmedia, for example, pegs it at 100%, and their chart shows it even exceeds that standard a bit on all the extremes except green-yellow.
Color accuracy
The accuracy of the color for the new MBP is measured by Notebookcheck at an average of 1.4 Delta-E 2000, with a maximum of 3. That's a measure of the difference between the displayed color and the color that's supposed to be displayed, so smaller is better. Originally a difference of 1 on this scale was supposed to represent the smallest difference the eye can detect by careful examination, but in practice it turns out that boundary varies from 1 to 2.5. At 1.4 the average accuracy is such that you would normally be unable to tell the difference. The maximum of 3 means that for some color(s) you probably can perceive a very small difference. Average Delta-E scores over 5 are common for laptop screens, so 1.4 is excellent. These are scores without calibration. With screen calibration (which requires tools that cost a little), you can get even better accuracy.
Last year's model scored a 2.1 average for Notebookcheck, also very good.
The XPS UHD was measured by Notebookcheck at 5.34 average, maximum 7.86, not so great. After calibration, it improved to 3.7 average, still not great. Anandtech, though, measured the screen at 3.19 average, and 1.0 after calibration, which is excellent. 4k.com measures it at 2.61 before calibration and 1.96 after. Laptopmedia measured an outstanding .52 average after calibration.
CPU
These are the average Geekbench 4 scores of the new MBP 15" CPUs. These may include measurements for different machines that use the same processors. The numbers can shift over time, I've noticed, and the scores don't represent how the processors do for every task, of course, so they're a rough guide at best. The first number for each processor is for single-core operations, the second for multi-core, followed by the percentage increases from the corresponding scores of the preceding model.
2.6 GHz i7-6700HQ: 3688 11389
2.7 GHz i7-6820HQ: 3999 12586 +8.4% +10.5%
2.9 GHz i7-6920HQ: 4291 13669 +7.3% +8.6%
https://browser.primatelabs.com/processor-benchmarks
GPU
The Geekbench 4 scores for the new MBP 15" GPUs, plus that of the 2015 and the XPS UHD. The same caveats as above apply.
Radeon R9 M370X (2015 MBP 15"): 37823
Radeon Pro 450: 41055 +8.5%
Radeon Pro 455: 47420 +15.5%
NVIDIA GeForce GTX 960M (XPS UHD): 47818 +0.8%
Radeon Pro 460: 52152 +9%, +10% over the 455
https://browser.primatelabs.com/opencl-benchmarks
*************************
Battery life
The best comparisons can be made with controlled tests by the same people. Comparisons to the XPS are complicated by the fact that early reviews were before software updates that reportedly improved battery life. The only uniform tests I could find for all three computers are from Ars Technica, whose review of the XPS was after the software updates. Here are their results:
New MBP
Wi-fi browsing, 200 nits: 15:33
WebGL, 200 nits: 2:49
Old MBP
Wi-fi browsing, 200 nits: 10:03
WebGL, 200 nits: 2:53
The new MBP lasted about 50% longer than the older MBP for the wi-fi browsing, but the battery life was similar for the heavy-duty WebGL benchmark.
XPS
Wi-fi browsing, 200 nits: 8:02
WebGL, 200 nits: 3:45
The new MBP lasted over 90% longer than the XPS in the wi-fi browsing, but the XPS lasted 33% longer for the more intense benchmark.
Notebookcheck tested the two MBPs:
New MBP
WiFi Surfing, v1.3, 140 nits: 10:02
Big Buck Bunny H. 264 1080p, 140 nits: 8:36
Load (max brightness): :55
Old MBP
WiFi Surfing v1.3, "about" 150 nits: 7:12
Big Buck Bunny H. 264 1080p, about 150 nits: 6:53
Load (max brightness): 1:45
The new MBP lasted 39% longer for their wi-fi test. For the video streaming test (60 fps, I believe), the new MPB ran 25% longer. (The possibility the old MBP screen was set a bit brighter could have made a small difference.) When they ran the machines all out on full brightness, the old MBP lasted 90% longer. Of course, the new MBP's screen gets about 50% brighter.
Engadget ran video playing tests on the the new MBP and the XPS and usually got 9 to 10 hours (up to 11) from the MBP at normal brightness. They don't say if they ran the test as many times with the XPS, which they listed at 5 hours 25 minutes (after the software updates).
All these tests suggest the new MBP outlasts the old MBP and the XPS for light and medium duty, but gives way to the larger batteries of the others for heavy-duty work.
RAM energy usage
The new MBP battery has a 76-watt-hour capacity. The largest battery the FAA allows on planes is under 100. The Dell XPS allows 32 GB RAM, but to do that with current laptop CPUs requires using desktop RAM, DDR4, instead of the LPDDR3 the MBP uses (LP for low power). I figure that costs about 5 to 84 watt-hours over ten hours for 32 GB, depending on how much you use the RAM, and 2 to 18 watt-hours for 16 GB.
If you really want to see the math behind that . . .
According to the chart at this link, the kind of RAM used by the new MBPs, LPDDR3, uses 70% of the energy used by regular DDR3 when active, and 10% when on standby, i.e. when powered but not active. (The LP stands for low power. It's designed for mobile devices.) Put the other way, DDR3 uses 43% more power than LPDDR3 when active (100/70 = 1.43) and 10 times more on standby.
https://blogs.synopsys.com/committedtomemory/2014/01/10/when-is-lpddr3-not-lpddr3-when-its-ddr3l/
The Dell comes with DDR4 2133 RAM. It's supposed to use less power than regular DDR3, but in practice people aren't seeing much difference.
http://www.tomshardware.com/reviews/asrock-fatal1ty-z170-gaming-k4-d3-ddr4-vs-ddr3,4431-2.html
On the other hand, Table 4 on page 6 at the next link claims 4% power saving with an 18% increase in throughput, which would represent a 23% increase in efficiency (for not quite equivalent RAM):
https://drive.google.com/file/d/0B-W2SEvdVPhtU2VXVnNfdFp5Rlk/view
I'm more inclined to accept the results from practical use, which show little difference, but I'll suppose a 10% saving for DDR4. That means DDR3 uses 11% more energy than DDR4 when active.
According to the above source (which is measuring eight 16 GB sticks), two 16 GB sticks of DDR4 2133, which is the standard RAM speed on both the Mac and the XPS, use 11 watts of energy when active (44/4).
According to the next link, two 8 GB sticks of DDR4 2133 use about 6 watts.
http://www.tomshardware.com/reviews/intel-core-i7-5960x-haswell-e-cpu,3918-13.html
So equivalent DDR3 would use 6.6 W.
Using the 70% figure from above, that means the LPDDR3 used by the MBP uses about 4.6 watts when active. That's a difference of 6.4 watts between 32 GB RAM on the Dell and 16 on the Mac (11 - 4.6).
More RAM energy use also causes proportionally more heat, which requires more fan use to get rid of it. The fans in laptops are typically 2-watt, and there are two of them, so they would max out at 4 watts. We're into the realm of guesswork here, but let's suppose as a worst case that extra use of the fans use up an extra 2 watts, plus the 6.4 = 8.4 watts more than what the Mac would use when active.
There are various kinds of "standby" and "idle," so it isn't clear if the figures in Table 4 referred to above and those in the table at the first link refer to the same thing. Supposing they do, when idle the DDR3 RAM at Table 4 uses 14% of the energy it uses when active. If the ratio is similar to the 8 GB sticks, two 8 GB sticks would idle at .9 W. And by the 10% figure from the first link, the equivalent LPDDR would idle at .09 W. According to Table 4, 32 GB of DDR4 2133 idles at .6 W, a difference of about .5 W.
So, the range of difference is .5 to 8.4 W between 32 GB DDR4 and 16 GB LPDDR3, depending on how much you use the RAM (and some guesswork).
Two sticks of 8 GB DDR4 use about 6 W, 1.4 more than the 4.6 the Mac's LPDDR3 uses. That's about 1/5 the difference with 32 GB, so the extra fan usage would be maybe .4 W more, for a total of 1.8 W.
By the figures at Table 4, the DDR4 idles at about 5.5% of active power. That would be .33 W for two 8 GB sticks. That's .24 more than the .09 W used by the Mac RAM when idling.
So the range for 16 GB is .24 to 1.8 W extra for the DDR4 RAM.
Over a period of 10 hours, 32 GB RAM DDR4 would use 5 to 84 more watt-hours.
Over 10 hours, 16 GB DDR4 would use 2.4 to 18 more watt-hours.
