Newsletter Issue #898
February 13th, 2017THIS WEEK’S TECH NIGHT OWL RADIO UPDATE
So we know there’s a fairly normal routine for Apple and Samsung and their flagship smartphones. While the iPhone SE, introduced last spring, was a notable exception, Apple normally introduces new models in early September, releasing them for sale a week or two later. The other assumption is the “tick-tock” upgrade cycle, where there’s a major form factor change one year, and a minor refresh, so to speak, the next. But the iPhone 7 is said to be an exception since it seemed so close ion appearance to the iPhone 6s.
For Samsung, the latest Galaxy lineup will usually debut at the Mobile World Congress in Barcelona, or at a separate event around the same timeframe, to be put on sale some weeks later. So there have been expectations about what the presumed Galaxy S8 would contain, except that it may not arrive when many expected. It appears that there will be no such product release for now, and it’s not at all certain when the new models will be launched.
One key reason is the failure of the Galaxy Note 7 phablet ,which had a nasty habit of overheating or bursting into flame due to a battery defect. Samsung is certainly taking a reasonable approach to slow down product development until they are certain the batteries in their mobile handsets are reliable. But it also means that Apple may have a huge leg up this year assuming the next iPhone arrives as scheduled later this year, without a competing smartphone from Samsung.
But imagine how it would be if that failure came from Apple, since so much of its sales are dependent on one product line? At least Samsung can continue to launch new washing machines that, one hopes, won’t fall apart under normal use.
Now on this weekend’s edition of The Tech Night Owl LIVE, we featured outspoken commentator Peter Cohen. During this segment, Peter went into detail on the value of the Apple Watch, primarily for fitness, and its other potential benefits. What about the iPad? Is there a way for Apple to persuade more people to buy them? Gene explained why the iPad doesn’t suit his workflow. There was also a long discussion about Apple’s use of an ARM processor and watchOS-style OS to power the Touch Bar on the new MacBook Pro, and the possibilities that Apple might be gradually moving the Mac away from Intel to ARM.
You also heard from columnist Derek Kessler, managing editor of Mobile Nations, a web portal that runs several popular tech sites.. In advance of the 2017 Mobile World Congress in Barcelona, Spain, Derek looked over some of the possible product announcements there. What about smartwatches? Does Apple have a clear path with the Apple Watch in light of the soft sales reported by Fitbit and the Samsung Galaxy Gear? How about the smartphone market? Are there compelling features manufacturers can add to entice customers to continue to upgrade every two years or so? Why did Samsung decide not to introduce new Galaxy smartphones at the Congress this year? What about the future of tablets? Are there killer features to boost sales of the iPad and other gear? Derek also explained why he thinks Apple is destined to move the Mac to ARM, and how that decision will really mess up Intel.
On this week’s episode of our other radio show, The Paracast: Micah Hanks, of The Gralien Report, returns to discuss strange objects in orbit around the Earth that were being seen during the golden age of Ufology. Researcher Donald E. Keyhoe believed some of these were “alien motherships”; Micah will look at how the discoverer of Pluto became involved in determining what these objects were, as well as the peculiar search for Earth’s second moon, and the continuing search for “moonlets” today. Over the years, mystery objects in Earth orbit have helped promote the belief in what is known as “The Black Knight Satellite,” a hypothetical alien probe that has monitored Earth for 13,000 years. Micah explains the scientific discoveries behind this modern legend, which have built onto a mythos over the years.
SO IS APPLE GETTING READY TO DITCH INTEL?
In the latter days of the PowerPC, development had clearly stalled. Although you could buy a Power Macintosh G5, a forerunner of the Mac Pro workstation, there was no similarly outfitted notebook. PowerBooks were still saddled with the G4 The reason? IBM and Motorola were not able to tame the beast to work within the constraints of a mobile device, or perhaps they didn’t care.
Indeed, some Power Mac G5 configurations required liquid cooling, and even with just fans, the system required several operating in an extremely sophisticated environment to keep the units from overheating. If the coolant leaked, your computer was toast.
At the time, Steve Jobs told interviewers that he was pleased with IBM’s PowerPC product roadmap, but that Apple was always considering its options. A key problem was the fact that other computer makers weren’t using PowerPC chips, which became more popular in the embedded market. So there wasn’t much incentive to deliver the parts Apple needed.
Jobs’ was prescient, reflecting plans that was already underway inside Apple. So during the 2005 WWDC keynote, he revealed that Apple would ditch the PowerPC and move to Intel processors the following year. With the arrival of the Mac Pro in the summer of 2006, the transition was completed months ahead of schedule.
That was then, this is now.
Mac users are complaining because product refreshes have stalled. Even when new models arrive, the updated Intel parts aren’t all that much faster than their predecessors. With notebooks dominating the PC market, Intel has focused very much on improving power efficiency rather than improve performance. Besides, most computers are perfectly adequate for routine tasks, such as email, web surfing and word processing, not to mention watching Netflix. So the incentive to fare better in benchmarks is less significant.
But Intel is also falling behind its annual schedules for processor upgrades. Apple was criticized for not using Kaby Lake silicon in the Late 2016 MacBook Pro, but the quad-core versions of those chips weren’t even shipping when these models arrived.
Since Apple is skilled at processor transitions, some suggest it should drop Intel and produce faster desktop versions of its ARM-based A-series processors. But even the parts they use, such as the A10 Fusion used in the latest version of the iPhone, are capable of PC-level benchmarks. And that’s for a processor optimized to do well within the severe constraints of a smartphone, where power efficiency has to be emphasized to the max.
Now most of you know that the newest MacBook Pros with Touch Bar include an ARM processor and an OS based on watchOS to power Apple’s OLED-based replacement for the venerable function keypad. The integration is, for all intents and purposes, seamless. If you didn’t know how it was done, you’d never notice.
According to published reports, Apple’s next target is the Power Nap feature of recent Macs, which can perform a number of background tasks while the computer is in Sleep mode, such as downloading updates and email. This would mean a more powerful processor and support OS, and that may only be the beginning. Well, if it happens of course.
So perhaps, as Apple removes some of the load off Intel silicon, it will mean more efficient power use and improved performance. I wouldn’t be in a position to suggest how much, but it may help Apple compensate for the lapses in Intel’s processors without actually doing a full processor switch. It may even be possible to follow the spirit of the dual-graphics scheme used in the 15-inch MacBook Pro. For normal functions, your Mac would run on ARM, using an advanced emulation system to accommodate Intel-based apps with minimal performance degradation. But when high CPU horsepower is required, and for running Boot Camp or a virtual machine, it would switch to Intel.
While a dual-CPU scheme of this sort may seem to be a rather awkward way to manage a processor transition — and it would certainly cause Intel conniptions — would it serve as a way station for a full processor switch?
Obviously, such a move would have its downsides. One, of course, is emulating Intel hardware in software, to allow you to continue to use your older apps. The A-series CPU would also have to be faster than Intel silicon to reduce or eliminate the possible performance loss. Developers shouldn’t have much difficulty making “fat” or “Universal” apps, since Apple perfected that technique long ago.
The real issue is the fact that many people require speedy Windows virtual machines or Boot Camp on Macs. Under Boot Camp, performance is identical to pure Windows hardware. Virtual machines come close enough for most anything but games, where they manage passable performance.
If Apple can devise a way to handle x64 instructions in its ARM hardware without a perceivable performance hit, however, a processor switch would be a no brainer. Or perhaps Intel will try to find ways to accommodate Apple’s needs going forward. While other companies buy more parts from Intel, Apple concentrates largely on CPUs that yield higher profits, and thus have more of an impact on the company’s bottom line.
So it’s possible the early use of an ARM processor on a Mac for the Touch Bar is not just a scheme to push the platform further ahead of Windows, but a warning shot to Intel to get its act together. Then again, the A-series processors cost Apple but a fraction of what Intel hardware costs. So an all-ARM Mac may sell for a significantly lower price, thus addressing a key objection by some to switching.
I can’t say how this will all play out in the end. As I said, there are tradeoffs in a processor switch, and it may not be worth the effort to make this move, except in limited ways.
THE FINAL WORD
The Tech Night Owl Newsletter is a weekly information service of Making The Impossible, Inc.
Publisher/Editor: Gene Steinberg
Managing Editor: Grayson Steinberg
Marketing and Public Relations: Barbara Kaplan
Sales and Marketing: Andy Schopick
Worldwide Licensing: Sharon Jarvis
Print This Issue
The SOC of a iPad Pro may be a reasonable replacement for a Mac mini, MacBook, MacBook Air or 13″ MacBook Pro which use integrated graphics. Do you see a future A series processor used in a 15″ MacBook Pro or a iMac?
If Apple plans to move Macs to ARM, there would be a family of CPUs that perform appropriate to the model.
Peace,
Gene
The 2016 MacBook Pros have the USB Type-C Thunderbolt 3 ports and Thunderbolt is from Intel. I don’t see an ARM-based Mac with Thunderbolt. It’s not likely that Apple would drop Thunderbolt. However, the MacBook could be a candidate for the processor transition. But, the roadmap of Intel is looking better for the MacBook due to the focus on power efficiency. I think the wholesale switch to ARM would be disrupted and could be costly in time and money. How does this moves the Mac platform forward?
I believe that Apple will stay with the x64 platform and continue to tweak the Macs with ARM.
Another intriguing idea–why not develop code for both?
My application needs to run on Intel, but why can’t I have certain threads that run on ARM? Apple certainly has ARM compilers. If there’s both Intel and ARM code, it could run on any CPU that happens to be available. Maybe even handle swapping data between the two.
For example, I have “background” tasks that need to be done–say, download something and interpret the response. I provide Intel and ARM code to do these tasks. I schedule it as a background task with GCD, so I’m not really worried about how long it takes. When it comes time to do the task, GCD looks at the current state of the machine and decides which CPU to schedule it on. If the machine is sleeping, for example, it might use the ARM processor because waking up the Intel CPU to do this will eat battery. If the machine is awake but idle, it could use the Intel processor. If the machine is awake and the Intel processors are busy doing calculations on some huge bitmap, it might use the ARM processor.
Heck, make it more entertaining–my download task gets scheduled on the ARM because the Intel CPUs are busy. Since I’m downloading, I’m waiting for data from the server. While I wait, an Intel CPU frees up. The data comes in and the analysis is done on the Intel CPU. So it mixes and matches depending on what’s available and how important the task is.
Some of these proposed two-processor schemes have a very Rube Goldberg sound to them. Does anybody have a handle on how large a group we’re talking about? What percentage of the Mac user base actually relies on Boot Camp? Large enough that it makes any sense (from either a business or an engineering point of view) to agonize about its welfare? I may be wrong here, but my sense is that the answer is a definite “no.”
Of course, if the question is how many of us rely on USB Type-C Thunderbolt 3 connectivity, the answer is almost everybody, and it would be colossal (and expensive) pain in the butt if we had to move on to something else. But if Apple opted for ARM chips is there any reason they couldn’t continue to use these ports under license from Intel?
If Apple went to ARM, obviously existing ports and macOS features would be supported.
Peace,
Gene
t was a good Monday to get to read two opinion articles about iPads. Read one from Rene Ritchie at iMore and still in middle of the other one from Daniel Eran Dilger from Appleinsider. Daniel’s editorials always get scores of comments that I will take the time to read. I just like reading reasonable arguments.
I see that there should be a “keep it simple” philosophy pertaining to all iOS devices and perhaps an advance mode to those who want to go beyond the basics. I recalled in classic Mac OS there was a Simple Finder where some of the features of the Finder were not present. Then, there was QuickTime, a basic media player where $30 would get you a license from Apple to unlock the “pro” features. Maybe there should be a “pro unlock” in iOS for those who want to do more on their iPad Pros.
On the aspect of iPad sales Apple should consider some suggestions like promotion as the one with videos shot with an iPhone or the AppleMasters for Macs. All iPads should have solid features like four speakers, True Tone display, etc. A way to get more new casual iPad users onboard is better pricing.
Please ignore due to inattentiveness when posting, wrong topic.