Just a heads up - Beats are crap
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Supply Chain Expects New 16-Inch MacBook Pro, 13-Inch MacBook Pro, and MacBook Air to Launch in October
- Thread starter MacRumors
- Start date
- Sort by reaction score
Just a heads up - Beats are crap
Apple could do it differently. They could put the cursor keys or "the middle part" of dekstop keyboard to left side and numbers to the right side. Or just try to find elegant lookin way to place the keyboard non symmetrically.
[doublepost=1563964668][/doublepost]
Could you live in a Universe, where Apple offered TWO choices: one overslimmed (with thermal and keyboard issues) and second fat model, which is heavier, has some other ports and slots than just usb-c and of course bigger battery. Maybe even 2nd internal storage (sd slot was this, ec slot before that...)?
Current chassis is 349 mm wide. Current screen is 333 mm wide. Going from 15.4" to 16.0" widens the screen 13mm. With current chassis that leaves 1.5 mm per side. Just by adding few millimeters to chassis size, it would fit nicely.
On the other dimension, if the Facetime cam is under the screen, they could reduce the depth greatly.
For usability, I guess they want to keep small black besel so it's easier to see where screen ends. Also, with some bezels, the screen does not break so easily...
Small gripe and it’s not just Apple. Having a 15in. (or 16in.) display and keyboard with no numeric keypad. A lot of people work with lots of numbers and having that numeric keypad built in is really convenient rather than have to use a plugin keypad. There is certainly enough real estate there. At that price-point the device should be more versatile.
...and about the rumored transition to ARM, first in laptops, I'm not so sure. I don't think it's a good idea, but Apple just might be crazy enough to do it. Two points:
1. Don't count Intel out. Their foundry might be suffering right now, but their design prowess has not failed. They might be stuck on a process twice the size of where TSMC is at, but the Core-series significantly out-performs the A-series (despite the significant gains made). If Intel overcomes their manufacturing issues, watch out.
2. At the office, we have 8 Macs. Only 3 of them run MacOS. The other 5 are running Windows or Linux. People don't always buy Apple hardware to run MacOS. Obviously we don't buy servers or workstations from Apple, cause they don't make any, but of the things they do make, they make really good ones, and the "Apple tax" is fiction. We know, we also buy product from HP, Dell, and others... Anyway, if Apple drops Intel, we'll probably have to drop Apple as a hardware supplier. It's not a criticism, it's just reality.
1. Don't count Intel out. Their foundry might be suffering right now, but their design prowess has not failed. They might be stuck on a process twice the size of where TSMC is at, but the Core-series significantly out-performs the A-series (despite the significant gains made). If Intel overcomes their manufacturing issues, watch out.
2. At the office, we have 8 Macs. Only 3 of them run MacOS. The other 5 are running Windows or Linux. People don't always buy Apple hardware to run MacOS. Obviously we don't buy servers or workstations from Apple, cause they don't make any, but of the things they do make, they make really good ones, and the "Apple tax" is fiction. We know, we also buy product from HP, Dell, and others... Anyway, if Apple drops Intel, we'll probably have to drop Apple as a hardware supplier. It's not a criticism, it's just reality.
Sure it has. They've been on the same process size for five years now, and have had to ship several minor revisions to the Skylake microarchitecture to make up for it (because they traditionally don't want to ship a new microarchitecture without first having changed the process). At some point, that isn't going to cut it any more.
TSMC's 7nm is more comparable to Intel's 10nm, actually.
And no, at similar TDP, Intel does not outperform Apple. Quite the opposite, in fact. Where Intel does outperform Apple is on the high end, where Apple doesn't actually have any comparable chips. We simply do not know what a 15W, 28W, 45W or higher Apple A-series chip would be like. But we do know that Apple A-series has been destroying Intel's 5W Y-series parts.
No, that's incorrect. What you're paying for is the base speed when idle. Turbo Boost is a freebie and not even guaranteed, otherwise, the CPU would be rated at the Turbo boost clock speed and not at the base clock speed. If Apple cannot maintain base clock speed of the CPUs in their computers under load, then that is a completely different issue and has to be rectified by Apple through better thermal design. But if you think Apple or any other PC OEM should be designing computers to maintain boost speeds all day, every day 24x7, you really aren't understanding how Intel's SpeedShift, Turbo Boost and Thermal Velocity Boost are designed to work.
Amen. A return to a good, quality keyboard would be an important step too.
At this point, hard to say. My speculation is that Apple (or Intel itself) isn't convinced that Ice Lake will ship in volume any time soon.
So it might indeed be that, as a result, they're pushing Comet Lake ahead of schedule. Which won't be very exciting. If the leaked S parts are accurate and H will be roughly the same, probably a move from i5 with 4 cores, i7 with 6 and i9 with 8 to 6 / 8 / 10 cores.
If you are running Windows or Linux on 5 of the the 8 Macs in your office, then Apple transitioning to ARM probably is a bad thing for you as it means you will have to move to another PC OEM (Acer, Dell, HP, Lenovo) for your Windows and Linux workstations or you will need to build them yourself.
I went through the 6x000 to PowerPC transition and the PowerPC to Intel transition and they had very few hiccups along the way, but it sounds like you're wedded to Intel, which is too bad, as they may make some very performant CPUs, but you are paying through the nose for the privilege while you watch them continue to market incremental steps as brand new "Gens" of CPUs. Unfortunately, 10nm is dead, Intel just refuses to call the time of death and its customers continue believing they can pull a rabbit out of their hat. To quote you - "If Intel overcomes their manufacturing issues, watch out." - this sentence has been the mantra of many a tech pundit for the last 3+ years and Intel has produced exactly one released 10nm CPU, the Core i3-8181U, and used that CPU in only a few systems sold overseas. It's a bust and Intel just needs to move on. They could have deflected some of this by simply integrating an LPDDR4/x memory controller into their 14nm CPUs, which would have made a world of difference for all PC OEMs in terms of battery life and memory expansion in their thin and light models. I reckon that Apple is tired of being held hostage to Intel's games and unlike Acer, Dell, HP, Lenovo, et al. they can and will make a change if they think it will give them more control over their hardware, which it will. I think they are taking a very measured approach with Catalyst and that there will be parallel A-Series and Intel computer lines to give more time for the transition than they have in the past. This is why there have been so many changes to macOS over the past three releases (High Sierra, Mojave and now Catalina) to pave the way for the transition and to give customers specific breakpoints. The recent MacBook Pro and iMac releases are proof that Apple is trying to give customers high performance systems that will last for several years as they make the transition. Perhaps they will never get rid of Intel-based systems, but I digress.
True. You wouldn’t want them operating at the boosted speed full time even if thermals weren’t an issue, as reliability is not rated under those assumptions - the increased currents and electric fields would accelerate electromigration, hot carrier degradation, dielectric breakdown, etc.
When we design the chip and simulate to make sure that it will have the required lifetime, we estimate how often each wire will switch, for example. Running the processor at the highest possible speed obviously increases the number of switches. Unidirectional current paths will then have faster electromigration failure than predicted. Same problem for other high field degradation effects. Higher average voltages than assumed, higher fields than assumed, etc.
Yup.
They bet big on bringing lots of advances with 10nm and then… couldn't ship 10nm. Then they doubled down on stubbornly not adding LPDDR4 to Whiskey Lake, Amber Lake, whatever.
It is hard to say, and I may be trying too hard to read the tea leaves and glean the future, so I readily admit that I am making an educated guess. I think I base my guesses on the whole Broadwell debacle and how much trouble they had with 14nm yields, even though they eventually found their mojo.
Intel has been a mess for a while organizationally, and really seem content to continue pumping out 14nm refinements, while simply increasing core counts to make up for the lack of any true gains in performance. I believe that they have pushed x86 as far as they can and still coming to terms with where to go next. In the meantime, AMD is moving ahead at a fair clip and if it can manage its growth along with the pains of upscaling their organization while not blowing the opportunity, things get very interesting for Intel for the next few years.
No, Comet Lake is not very interesting, but I don't have very high expectations for Intel anymore. There are simply better companies and CPU that have come and gone, Intel seems to get by on luck and sheer size. They don't call it Chipzilla for nothing. Only time will tell and Intel is a survivor, if no longer an innovator.
And don't forget these machines will probably not be able to run any 32 bit apps.
And that stubbornness seems more like desperation as they continue pumping out those 14nm refinements, without adding the LPPDR4/x controller. Apple has been using LPDDR4 or 4/x in their iOS devices since 2015.
This is actually an interesting point. Does Intel guarantee that Turbo Boost will not reduce the chip's lifetime?
Your numbers are pretty close (I get 332 mm screen width based on both measurement and calculation), but you're forgetting the case and gasket thickness (and since you're calculating to the mm, you need to take these into account): Take a look at the L or R edge of the display. You'll see a thin line of silver, from the metal case, and then a thin black rubber gasket. These take up ~2 mm on each side. So with a 349 mm wide chassis, you have ~345 mm for the screen, i.e., no extra space on the sides at all if you go to a 16.0" diagonal (332 mm + 13 mm = 345 mm). And I don't know if Apple will go right to the edge, so 16.0" is a theoretical upper limit on the display diagonal (assuming they keep 16:10) in the current chassis, not what they could install in practice.
This is a really confusing time to be in the market for a 13" mac product. Too many choices/updates and hard to figure out if one is really getting value. Part of me wants to just buy a base machine with the assumption it will be replaced soon when the dust settles (e.g. keyboard issues, DDR4, etc.).
It will be interesting to see whether the dongle & connector situation improves any time in the next year. It isn't just the cost; it's the overall quality of available docks, port replicators and connectors. USB-C ports that don't support Thunderbolt, USB ports that don't provide enough power, video connectors that don't include VGA (sorry, but I present at venues worldwide and about half still require VGA), DVI connectors that only work with particular adapters in addition, HDMI that doesn't support the 2.2 standard, and on and on. Add $800 to the cost of a USB-C-only device if you want to do serious work on these things.