No matter what else, I'd like to see 10-hour battery life and laptops that run cool! The MacBook Pro is more than fast enough for many, if not most users, but they run hot and cannot match the MacBook Air's battery life.
Most of all, i want to see SSD storage options priced in line with 2016 expectations.
(I need 1TB internal storage.)
I need 2TB or I'm likely going to be waiting another year if I don't do some serious pruning, which I don't feel like I should have to do. I have 2TB in my 2011 right now.
If it's true that Kaby Lake doesn't require an Alpine Ridge controller for Thunderbolt 3, that's very interesting for a 12" MacBook update and possibly the MacBook Air -- which could happen any day now.
As for the wildly anticipated MacBook Pro, just upgrading them to Skylake this year is a nice improvement, especially the 15" which is still rocking Haswell. There should be ample space for Alpine Ridge to still have Thunderbolt 3. Other than a slight ~10% boost, Kaby Lake's main improvements are in the fixed-function encode/decode hardware on the GPU. Not a big loss if there is an AMD Polaris GPU onboard with the same functionality.
The slow trickle of Intel CPUs certainly complicates matters for Apple's engineers, but I'm confident in their abilities to bring out a great new Mac lineup.
Considering Polaris wipes the floor with any iGPU from Intel, nothing Kaby Lake does with OpenCL/Metal API would touch an RX 480 mobile or custom ASIC design.
The changes in Mac hardware began under Steve. I think most people here have read enough about him to know he was not sentimental about Apple products. If something had served it's function, made the profits and was no longer valid he would dump it. I think he was headed that direction with Macs. It was under Steve that the Mac Pros languished almost unchanged for years. Steve killed the XServe and Xserve RAID which also had declined under his tenure. I think the reason they languished is that it was easy for him to see that things were changing in that market segment like Wigby said above.
according this article and quote, Jobs' reason for canning the server was much simpler -- "nobody was buying them".
They didn't sell because they were half heartedly supported and marketed enough so that only Mac faithfuls were in the least interested, and even then. I would say say the servers could have been great if done well but they never were going to fit under Apple's hardware manufacture, release and promotion philosophy. It's a different market, similar to pro audio and pro imaging, and Jobs was being forward thinking to when he didn't want to be the tail wagged by those dogs.
so on the one hand we have an armchair CEO who says they only didn't sell because Apple wasn't trying hard enough. on the other hand we have Steve Jobs, an actual CEO for the actual company selling them who felt they didn't sell because there wasn't a market. hmm. yeah, I'm gonna go with Jobs' take.
Did you even bother to read the linked article? Jobs says nothing about there not being a market for servers. He's quoted in the article as very plainly (allegedly) stating that the reason they were discontinuing the Xserve line was because "Hardly anyone was buying therm". Nothing about the server market. "Hardly anyone" bought into them during their run was because Apple put out a cool, sexy rack mount unit that IT people weren't impressed by. It was one of those ventures that you could feel the end result at the initial release. It was not a failure because the market was too dumb to know what to do with it. It was killed off because it was a misstep to think existing server users would fall in love with it and Apple never succeeded in selling it to Apple users.
And as far as me second guessing Jobs, nothing I said meant that I'm a genius and Jobs had no idea. You do realize that this is an Apple forum where half of the posts are armchair quarterbacking? : ) I even soft pedaled it by saying "I would say". You don't want to have to cut out every other post because someone has an opinion.
Whatever Apple chooses in the short term, intel is all but to the curb for future Macs. First Apple can't rely on whatever whims Intel imposes. Secondly, Intel has said publicly that the x86 is not a priority anymore with the death of the PC. Third, and most importantly, Apple has spent a lot of money on the A series. And the A series is not going to sit still. It's constantly improving. Faster I'd say than the X86 ever did. They have way too much invested in the A series to just follow Intel off the cliff they are marching towards.
No no, people keep drumming out this "Apple will use it's own chips" every year, but it's even more of a joke than using Intel's iGPU itself.
Apple, could, if it wanted to completely alienate it's existing userbase. This isn't like with OSX 10.3/10.4/10.5 where they were still in the late stages of migrating people off the PPC MacOS System 6/7/8/9 systems, and those installations were being thrown to the curb either way.
Such is not the same now. Apple will continue to use the Intel chips as long as they remain cost-effective for the performance. We hit a performance wall years ago, and every laptop produced since the beginning of the i5/i7 series has effectively the same performance, and only increases in battery life. The iGPU's are terrible, and always have been terrible and 3 generations behind AMD. Yet what does Intel do? Dedicate 60% of the die space to it.
So what does the A-series offer? Nothing. The A9 series is barely more powerful than the same kind of CPU/GPU from Intel that goes into it's ultra-low-power chips, not it's desktop chips. It has to meet or exceed the performance in the desktop chip before the A series is any kind of win, and it's the only fire lit under Intel's ass. You can be sure that if it wasn't Apple asking for high performance laptop parts, AMD would be more than happy to use it's far-superior iGPU solutions with it's weaker CPU solutions.
Apple's best use of the A-series is for a "thinner" ipad-that-has-the-performance-of-a-laptop and then start building "ipad" laptops rather than trying to build laptops with A-series parts. That will ensure that people stay locked into the Apple ecosystem, and aren't forced to dump their entire software library.
Maybe Apple should just license OS X it does run much better on other peoples hardware.
Although the current Mac Pro is getting a little long in the tooth - it is still more expandable than the old Mac Pro by far (I have the 2008 Mac Pro). I currently have two 5770 graphics cards sitting in the case and it leaves one smaller / lower bandwidth PCIe slot which is taken up by a SAS controller to an external hard drive chassis (at lower bandwidth than I would like).... and it was full.
Time to ditch the 5770s as the new Nvidia/AMD cards now have lower power draw. Check out www.barefeats.com to see what's supported.
Holy crap! Zen engineering samples were demonstrated to beat Broadwell-E with Blender running straight CPU to CPU [each benched at 3Ghz] sampling and rendering. Add in Apple and AMD's OpenCL stack, throw in the APU that is coming out for Zen and not a single person in these forums would be asking Apple to use Intel. Most certainly they would be asking to use the ARM A-Series chips that Keller pioneered for Apple, who now is the pioneer behind AMD Zen.
Apple needs to cut ties with Thunderbolt as Intel refuses to license this to AMD.
AMD Zen has USB-C 3.2 version 2.0 with 10Mbps throughput built-in along with every single other current tech that Intel has on their CPUs.
The cat's out of the bag for all it offers. Go see the specs on AMD's site.
There is no other compelling reason for Apple to use Intel exclusively than Thunderbolt. Thunderbolt is also the most overly hyped and priced serial port every invented. 99.9% of all consumers never touch it.
Apple has custom ASIC designs from AMD coming very soon. If they would pull the trigger on FX and APUs from AMD costs for Apple systems would be considerably lower and performance would actually improve.
AMD Zen chips are committed to 16 and 32 physical cores with 32/64 threads respectively. Apple could choose to adopt the Summit Ridge 8 physical core/16 threads for iMac and the beefy 16/32 Naple cores for Mac Pros if they wanted. They can have custom ASIC designs made from AMD and manufactured by Samsung/Global Foundries and even TSMC.
All this may be true. I have no reason to doubt it.
Id suggest Apple wait a couple years to see if AMD can keep it up.
I'm not so quick to dismiss Thunderbolt. It is certainly my preferred way to move multi-gigabyte projects around. I like the port consolidation as well.
As I mentioned in another post, there is more than pure computational performance that makes a 'pro' level Mac. I like the port consolidation. I think the soon-to-be reality of one type of port that covers everything from keyboard input to 4K display is brilliant. This will simplify the design of the new MBP significantly, allowing fewer, smaller ports to do the job that used to require many more, bigger ports.
I'm not so quick to dismiss Thunderbolt. It is certainly my preferred way to move multi-gigabyte projects around. I like the port consolidation as well.
I like thunderbolt and think it is a great.... just wish it were an open standard.
The port consolidation is a USB standard which Thunderbolt has more or less added on to.
A vendor could also extend the USB-C standard and use the alternate data path to define other standards other than HDMI (new), Displayport, Thunderbolt and USB.
I would think any serialized data standard could be created to add on like an USB-C:SAS, USB-C:MIDI, USB-C:Ethernet or other low latency alternative that could give you what you have in a unified USB-C / Thunderbolt standard but at a lower level - and better real-time transmission if it were important to a vendor. Thunderbolt is advantageous for an external bus for add on cards, but for the most part it tends to be used for storage, display and expansion.... something that would not be as much of a need if you had those supported sort of natively.
Most of all, i want to see SSD storage options priced in line with 2016 expectations.
(I need 1TB internal storage.)
Since when has Apple priced storage options inline with current expectations?
Your 'expectation' is unrealistic.
Actually, their prices for PCIe SSDs were pretty okay when they came out, IIRC. Much more expensive than "regular" S-ATA SSDs, of course, but much faster, too.
I'm not so quick to dismiss Thunderbolt. It is certainly my preferred way to move multi-gigabyte projects around. I like the port consolidation as well.
I like thunderbolt and think it is a great.... just wish it were an open standard.
The port consolidation is a USB standard which Thunderbolt has more or less added on to.
A vendor could also extend the USB-C standard and use the alternate data path to define other standards other than HDMI (new), Displayport, Thunderbolt and USB.
I would think any serialized data standard could be created to add on like an USB-C:SAS, USB-C:MIDI, USB-C:Ethernet or other low latency alternative that could give you what you have in a unified USB-C / Thunderbolt standard but at a lower level - and better real-time transmission if it were important to a vendor. Thunderbolt is advantageous for an external bus for add on cards, but for the most part it tends to be used for storage, display and expansion.... something that would not be as much of a need if you had those supported sort of natively.
Forgive me, I'm just a simple graphic designer, so I'm not sure I understand the difference you are drawing between what is predicted (USB-C carrying Thunderbolt as well as other adaptations) and the open solution you propose.
All I'm saying, like a nerdy Paul Simon born 40 years later may have sung in 2016: "So mama don't take my Thunderbolt away"
I'm totally on board with USB-C with Thunderbolt along for the ride.
I'm not so quick to dismiss Thunderbolt. It is certainly my preferred way to move multi-gigabyte projects around. I like the port consolidation as well.
I like thunderbolt and think it is a great.... just wish it were an open standard.
The port consolidation is a USB standard which Thunderbolt has more or less added on to.
A vendor could also extend the USB-C standard and use the alternate data path to define other standards other than HDMI (new), Displayport, Thunderbolt and USB.
I would think any serialized data standard could be created to add on like an USB-C:SAS, USB-C:MIDI, USB-C:Ethernet or other low latency alternative that could give you what you have in a unified USB-C / Thunderbolt standard but at a lower level - and better real-time transmission if it were important to a vendor. Thunderbolt is advantageous for an external bus for add on cards, but for the most part it tends to be used for storage, display and expansion.... something that would not be as much of a need if you had those supported sort of natively.
Forgive me, I'm just a simple graphic designer, so I'm not sure I understand the difference you are drawing between what is predicted (USB-C carrying Thunderbolt as well as other adaptations) and the open solution you propose.
All I'm saying, like a nerdy Paul Simon born 40 years later may have sung in 2016: "So mama don't take my Thunderbolt away"
I'm totally on board with USB-C with Thunderbolt along for the ride.
And if that means Intel-only, so be it.
What I was saying is if there are non-Intel Macs in the future, there will be no Thunderbolt.... If Intel stays as the only Mac alternative, then no need to necessarily expand it.
USB standard works well for low cost -- devices where latency is not the highest priority.
Thunderbolt is only used when the first is not the best alternative (since it costs more). (Side Note: Although Thunderbolt is lower latency than USB-C it still encapsulates PCI Express protocols -- as I understand it). Thunderbolt therefore will only generally get used for storage, docks, high-speed networking -- things that latency is important for.... for everything else... USB is a better cost/performance choice.
The USB-C standard is open, Thunderbolt is proprietary. I am not saying that it would be likely to implement other protocols, but it is not out of the realm of possibility. Intel is stumbling badly these days, leaving potential for other alternatives in the future.... whether it is ARM/OpenPower on server products, or ARM based consumer products... the USB-C standard is extendable.
Comments
(I need 1TB internal storage.)
Your 'expectation' is unrealistic.
And as far as me second guessing Jobs, nothing I said meant that I'm a genius and Jobs had no idea. You do realize that this is an Apple forum where half of the posts are armchair quarterbacking? : ) I even soft pedaled it by saying "I would say". You don't want to have to cut out every other post because someone has an opinion.
Apple, could, if it wanted to completely alienate it's existing userbase. This isn't like with OSX 10.3/10.4/10.5 where they were still in the late stages of migrating people off the PPC MacOS System 6/7/8/9 systems, and those installations were being thrown to the curb either way.
Such is not the same now. Apple will continue to use the Intel chips as long as they remain cost-effective for the performance. We hit a performance wall years ago, and every laptop produced since the beginning of the i5/i7 series has effectively the same performance, and only increases in battery life. The iGPU's are terrible, and always have been terrible and 3 generations behind AMD. Yet what does Intel do? Dedicate 60% of the die space to it.
So what does the A-series offer? Nothing. The A9 series is barely more powerful than the same kind of CPU/GPU from Intel that goes into it's ultra-low-power chips, not it's desktop chips. It has to meet or exceed the performance in the desktop chip before the A series is any kind of win, and it's the only fire lit under Intel's ass. You can be sure that if it wasn't Apple asking for high performance laptop parts, AMD would be more than happy to use it's far-superior iGPU solutions with it's weaker CPU solutions.
Apple's best use of the A-series is for a "thinner" ipad-that-has-the-performance-of-a-laptop and then start building "ipad" laptops rather than trying to build laptops with A-series parts. That will ensure that people stay locked into the Apple ecosystem, and aren't forced to dump their entire software library.
Id suggest Apple wait a couple years to see if AMD can keep it up.
I'm not so quick to dismiss Thunderbolt. It is certainly my preferred way to move multi-gigabyte projects around. I like the port consolidation as well.
As I mentioned in another post, there is more than pure computational performance that makes a 'pro' level Mac. I like the port consolidation. I think the soon-to-be reality of one type of port that covers everything from keyboard input to 4K display is brilliant. This will simplify the design of the new MBP significantly, allowing fewer, smaller ports to do the job that used to require many more, bigger ports.
I like thunderbolt and think it is a great.... just wish it were an open standard.
The port consolidation is a USB standard which Thunderbolt has more or less added on to.
A vendor could also extend the USB-C standard and use the alternate data path to define other standards other than HDMI (new), Displayport, Thunderbolt and USB.
I would think any serialized data standard could be created to add on like an USB-C:SAS, USB-C:MIDI, USB-C:Ethernet or other low latency alternative that could give you what you have in a unified USB-C / Thunderbolt standard but at a lower level - and better real-time transmission if it were important to a vendor. Thunderbolt is advantageous for an external bus for add on cards, but for the most part it tends to be used for storage, display and expansion.... something that would not be as much of a need if you had those supported sort of natively.
Actually, their prices for PCIe SSDs were pretty okay when they came out, IIRC. Much more expensive than "regular" S-ATA SSDs, of course, but much faster, too.
All I'm saying, like a nerdy Paul Simon born 40 years later may have sung in 2016: "So mama don't take my Thunderbolt away"
I'm totally on board with USB-C with Thunderbolt along for the ride.
And if that means Intel-only, so be it.
USB standard works well for low cost -- devices where latency is not the highest priority.
Thunderbolt is only used when the first is not the best alternative (since it costs more). (Side Note: Although Thunderbolt is lower latency than USB-C it still encapsulates PCI Express protocols -- as I understand it). Thunderbolt therefore will only generally get used for storage, docks, high-speed networking -- things that latency is important for.... for everything else... USB is a better cost/performance choice.
The USB-C standard is open, Thunderbolt is proprietary. I am not saying that it would be likely to implement other protocols, but it is not out of the realm of possibility. Intel is stumbling badly these days, leaving potential for other alternatives in the future.... whether it is ARM/OpenPower on server products, or ARM based consumer products... the USB-C standard is extendable.