Low-end Intel Kaby Lake processors detailed, MacBook Pro version absent
Intel has shed some specific details for the first time about the new seventh generation Kaby Lake processors suitable for the MacBook and MacBook Air, but as expected, processors suitable for the MacBook Pro have yet to materialize.
Corporate Vice President Navin Shenoy unveiled the technical information for two of the five new Intel Core processors on Tuesday. Intel's U-series is aimed at MacBook Air-like laptops, and the Y-series is intended for slightly higher-end devices like the Retina MacBook and PC convertible laptops.
Intel claims that overall the Kaby Lake architecture gives up to 19 percent better Web performance over similar sixth generation Skylake processors, and 12 percent faster general performance gains. Video processing of 4K streams is said to be over 6 times faster than a computer from five years ago.
Also included with Kaby Lake processors are power efficiency improvements. Intel claims up to 9.5 hours of 4K video playback is possible from laptops shipping in the fall with the new processor.
While the "GT2" integrated graphics in the Kaby Lake U-series processors are capable of better performance than existing processors for Apple's MacBook Air, Apple has historically used slightly higher end graphics in the line. It has been some time since the last product update, with the last revision dating back to March 2015.
The Y-series processor could ultimately find its way into the Retina MacBook, and would be an improvement in every regard over the Skylake processor currently found in the model. However, the Retina MacBook is Apple's most recently updated macOS product, and a refresh less than half a year after the last revision is not expected so rapidly.
The pair of Kaby Lake S-series processors suitable for iMac, MacBook Pro, and other higher-end deployments have been teased by Intel, but not yet detailed. Wide release of the S-series isn't expected any sooner than the beginning of 2017.
Intel revealed in late July that Kaby Lake processors had just started arriving in volume to unspecified manufacturers.
Corporate Vice President Navin Shenoy unveiled the technical information for two of the five new Intel Core processors on Tuesday. Intel's U-series is aimed at MacBook Air-like laptops, and the Y-series is intended for slightly higher-end devices like the Retina MacBook and PC convertible laptops.
Intel claims that overall the Kaby Lake architecture gives up to 19 percent better Web performance over similar sixth generation Skylake processors, and 12 percent faster general performance gains. Video processing of 4K streams is said to be over 6 times faster than a computer from five years ago.
Also included with Kaby Lake processors are power efficiency improvements. Intel claims up to 9.5 hours of 4K video playback is possible from laptops shipping in the fall with the new processor.
While the "GT2" integrated graphics in the Kaby Lake U-series processors are capable of better performance than existing processors for Apple's MacBook Air, Apple has historically used slightly higher end graphics in the line. It has been some time since the last product update, with the last revision dating back to March 2015.
The Y-series processor could ultimately find its way into the Retina MacBook, and would be an improvement in every regard over the Skylake processor currently found in the model. However, the Retina MacBook is Apple's most recently updated macOS product, and a refresh less than half a year after the last revision is not expected so rapidly.
The pair of Kaby Lake S-series processors suitable for iMac, MacBook Pro, and other higher-end deployments have been teased by Intel, but not yet detailed. Wide release of the S-series isn't expected any sooner than the beginning of 2017.
Intel revealed in late July that Kaby Lake processors had just started arriving in volume to unspecified manufacturers.
Comments
BTW people this is exactly why Apple hasn't updated their Mac lineup. There's nothing significant to upgrade to! What do you expect Apple to do if nothing is available that makes a difference? Would a 200 MHz speed bump satisfy you just for the sake of Apple releasing updated products? This has nothing to do with Tim Cook, and everything to do with the fact that there isn't anything to update to. And to me, it makes no sense for Apple to release a totally new iMac or MacBook Pro with last years technology inside it which would basically be what they'd have to do if they were to release any significant updates to their Mac lineup.
I mean, I get it. Intel is just stretching out “Moore’s Law” as long as possible so they have more time to come up with something to increase performance once Tigerlake is eventually released (in 2020, if they keep pushing everything else back). But it’s causing everyone to suffer for it.
Why can’t things just be like the good old days.
2) Note, I stated within 5 years. I wouldn't be surprised if we see something within a year. The technology is there, but there are other, more logistical, considerations to allow for this to start happening with Apple's low-end "PCs."
1. AMD's Zen will be ballpark competitive with Intel's Core and AMD is actively marketing themselves as a custom x86 SOC producer. Apple could perhaps have AMD make custom SOCs for the Mac that would include Apple-specific bits (like the secure enclave for TouchID) along with an AMD GPU (not as good as Nvidia, but certainly better than what Intel offers) and the perfectly cromulent Zen. All for a much lower price than what Intel charges.
2. By switching to ARM, Apple would lose some customers who need x86 compatibility to run legacy Windows apps. I don't know how many customers we're talking about. My guess is maybe 5% or less of the Mac installed base. Not huge, but not totally ignorable -- there would have to be a benefit to offer this cost.
3. In order for a custom ARM SOC to make sense, it needs to beat a hypothetical Zen SOC in price/performance and performance/watt. And the margin of victory needs to be big enough to attract enough new customers to offset the loss from issue #2. In theory I think Apple could do that. But in practice I'm skeptical that Apple will put forth the effort to do it. No doubt they can afford it financially.... but for reasons that I cannot really fathom, Apple's enthusiasm for pushing the Mac platform forward just doesn't seem to be there. This disinterested attitude about the Mac makes no sense to me, but it sure does seem to be Apple's attitude nonetheless.
I'm confused as to why this is not reported at all despite AI previously reporting these would be the ones to watch for. All signs point to Kaby Lake chips suitable for the MBP later in 2017.
Everyone here predicting a move to ARM architecture for Macs... You realize this would require some kind of Intel architecture compatibility mode transition or it would kill the Mac developer community and Mac software library, right? Switching to Intel killed lots of good software that never got ports (because it upset the developer's income or the developer lost interest in the product enough that rewriting it for a new architecture was out of their scope of motivation).
It DID improve the likelihood of ports from Windows, but it took six years of frustrating transition, WITH a PPC emulation layer present to fill in the gap left by big monolithic developers refusing to change with the systems they sell software on (looking at you, Adobe). I'm not sure, but can ARM even provide usable performance on an Intel emulation layer?
What's the real problem here? Faster CPUs? I don't think so. It's the overall system architecture that people want improved, right? Intel has badly lead thunderbolt and almost ignored Apple's lead on high-PPI display tech (how many Windows machines can you buy with retina-equivalent displays? I know this stuff matters not at all to tech geeks, but it matters to content creators). Apple should be building their own support chipsets for these things to work with existing Intel CPUs so they can move on to DisplayPort 1.4 and standalone retina displays for Mac mini and Mac Pro. They already were forced to create their own display driver chip for the retina iMac, so why not keep going?
Intel has zero motive to get the current standards into their slowed release cycle. In fact, it's in the Intel shareholders' best interests to drag out product releases as long as possible and with as few changes as possible.
Bundling so much of the computer into one chip (the so-called "application processor") works in their interest and against the interests of customers like Apple who still try to distinguish their hardware from other builders (because, as already noted, the PC world doesn't care about things like transitioning to high-PPI).
Intel reached the end of what can currently be done with rapid upgrading and probably has to invest much more into development of each new step (if they can even physically accomplish any further gains with shrinking this architecture). Shareholders don't like that. Capitalism never was fond of physics (or reality).
Moving to another architecture would be mostly transparent for the average consumer of Apple products (so long as Apple gets ALL of their own software ported by day one, and Microsoft has a version of Word and Excel), but it will absolutely suck for 3rd-party developers and any professional users (content creators relying on third party music and art tools currently available on Apple's Intel platforms).
The transition would probably be worse than the PPC-to-Intel transition. Since Apple doesn't give a damn about the professional market anymore, I don't see this as a problem that would stop Apple doing it. Apple is still arrogant but less sensible about their design choices than they were a few years back. This change could be the final "Apple is doomed", but only for those of us that prefer to do their content creation on Apple hardware with Mac OS.
iOS users will be fine (and no, iOS isn't ready to replace all the professional tools; the music side has only just started to look like it might reach near-parity in a few more years). Hell, I'm STILL waiting for Apple to fix text selection & autocorrect in text views on Safari. Is that too "pro" a function for them to care about??
Less reliance on Intel is the key, but abandoning an entire architecture is a myopic control freak "all or nothing" reaction. Relying on monolithic architectures isn't the solution to anything but intel's desire to sell everything to everyone, but leaving it entirely is another type of self-inflicted injury.
Like Apple building their own chips to make a retina iMac possible, we need the business to go back to custom chips for unique improvements and dedicated processors for specific tasks, instead of a CPU doing everything and one company being the only provider of chipsets. We need competition again.
Where is AMD? They still make desktop and server CPUs, right? What's their DisplayPort support level?
The games industry gave us the GPU as one third of the processor trinity needed. When will a computer maker team up with an audio hardware company to include an audio-dedicated DSP chipset in this architecture as standard equipment? Then the CPU can be left to power the OS and support functions of general computing, relieving the need for faster CPUs every year (which seem unlikely to arrive any more).
Intel isn't going to make such a partnership. Apple could and should. They could convert the native VST and AU plugin business into a DSP-run plugin business. Current music software industry developers selling DSP solutions are fewer than they used to be, but they have a strong following among their users. Buying out the UAD2 DSP & plugin business from Universal Audio, for example, is within Apple's financial ability. Integrating the DSPs with Mac hardware and adding the UA-branded plugins into Logic would strengthen Apple in the pro audio world. It would go hand-in-hand with the notion of Apple being a one-stop-shop for music listening AND creation.
These DSPs could be used to boost current Apple initiatives like Siri & speech to text, and provide local processing for other such audio analysis products like music identification.
But this would be an expensive and risky investment that would either succeed and change the whole computer industry for decades to come, or fail and lose millions. Apple doesn't really seem to care about the computer industry anymore. They're going to force people to redefine what computers are so Apple can exit one computer business and dominate a new one. I don't disagree that computers need to be redefined, but they can't be redefined as "consumer gadgets only". The transition of desktops to tablets can never be 100% across all use cases.
Ok, even with the iOS business they still need to free the CPU from being in charge of everything. CPU for operating system glue and applications. GPU for visuals. A dedicated DSP for audio (and likely more). In fact, a dedicated DSP could be even more important for iOS devices than desktops because these devices can never support CPUs with the raw processing power of desktop workstations used by big content creation shops. iOS devices have a much nearer end-point on raw computing power growth.
...but today's Apple won't do things like this. They're stuck in a rut of dragging out as much profit with as little investment as possible. Just like Intel. I don't know how the board was convinced to let Apple spend billions on building a car (I'm still hoping this rumor is absolutely wrong), but these same people see no value in professional users and content creators... mostly because they're not creators themselves. Typical myopic "one size fits all" corporate thinking.
It wouldn't surprise me if Intel themselves added some kind of DSP to their future chipsets or the CPU package itself, and marketed it as "the next step in PC performance for musicians and studios". Once they can no longer sell CPUs that make big sales, they'll have to offer something else to this market. GPU as DSP hasn't caught on and it shouldn't. Attempts at it have failed because the GPU isn't suited to it, and it needs to do its primary task of running the graphics and display.
Apple ought to get there first.