Patty’s Pioneers’ Predictions for 2011
By Scott Jordan, One of Patty’s Pioneers, January 6, 2011
NEXT GEN ITUNES: SOMETHING WICKED-COOL THIS WAY COMES
The Achilles’ Heel of the iExperience is iTunes, which is showing its age and creaking under the burden of doing stuff it was never intended to do: managing firmware and baseband updates, streaming videos, syncing contacts, apps, etc., etc. iTunes started as a portal for content for personal computers and iPods. It wasn’t very good even then. Now the storefront of the world’s largest music purveyor, it’s being asked to do too much, and it’s never escaped its casting as Apple’s most un-Apple product.
The need to hook up your pocket pal or tablet to a personal computer—in fact, the need to even OWN a personal computer—may be the most glaring shortcoming for the user experience in iDevices. I highly suspect that Apple is well aware of this and has been carefully watching the fledgling attempts by Google, some carriers, and the jailbreak community to facilitate wireless syncing. But Apple wouldn’t be satisfied with JUST wireless syncing—they would want such a feature to provide a real “wow” for their customers.
Meanwhile, recall that Steve Jobs called the iPad the most important thing he’d ever done. I don’t think he was referring to it as just a slim new form-factor or just a fresh new user interface; I suspect he views the iPad as the tip of an emerging iceberg of new ways of relating to our digital lives.
So my top prediction is for a sweeping revamp of the computing experience—yes, another—starting in 2011 with a generation of iOS that is at least optionally divorced from “real” PCs, fed by an iTunes-in-the-cloud that offers a content content—and computing—experience fully separated (if the user desires) from a hard-disk-based store for one’s media and documents. The current Apple TV gives a peek at how parts of this might work, but my iVision amps up considerably from that on the innovation-o-meter, making good on the idea that one’s computing profile and personal preferences and settings travel with you in your pocket, coming to life whenever you sit down at a workstation. Apple’s aim will be no less than to drive another revolution in personal computing, integrating the best of what we have today with the best of what cloud computing promises. The pieces are coming into place in 2011... just as the first-generation of competitors to the iPad poke their noses out of the nest. They’ll get stomped, because this isn’t about just-the-hardware, even now.
Apple’s Chip Investments: The Line in the Sand with Intel Is Philosophical
Three of the more intriguing tech events in the past couple years are, in no particular order:
(1) Apple’s purchase of PA Semi and Intrinsity.
(2) Apple’s evident turn away from partner Intel amid that company’s attempt to achieve relevance in the mobile market and its heavy-handed legal wrangling regarding nVidia’s GPUs (Graphics Processing Units).
(3) Intel’s purchase of McAfee, with its hints that anti-malware technology will be driven down into the chip level before long.
To my eye, the Intel/McAfee security-on-chip notion has always been about speed and, especially, power consumption. Just as video decoding is vastly more power-efficient when performed hard-coded rather than in software, similar savings could be achieved by baking security technologies into system-on-chip architectures, just as floating-point processors have long been integrated and GPUs are increasingly so.
To put this in perspective, consider the following table of percentage—slower horrors from a 2006 study of various applications when installed on Windows: the worst offenders are all security apps, some of which (including McAfee!) can nearly double boot times and slow file access by a factor of more than twenty: http://thepcspy.com/read/what_really_slows_windows_down/.
Intel, meanwhile, seems to be committed to the wide-open model where users can download and install anything from anywhere. After all, if Intel is the sole purveyor of microprocessors with hard-coded, GPU-esque anti-malware functionality, then bring on the malware! Every virus, every trojan, every worm becomes another reason to look for the Intel Inside sticker, in this view.
Clearly, anything a processor manufacturer can do to mitigate the burdens of always-running security processes will give it a significant performance and efficiency advantage and hence a competitive advantage.
...As long as such processes are necessary, of course. But as long as they are (and I’m increasingly feeling that’s a design choice), there are few conceivable aspects of computing which offer such tantalizingly high leverage for performance and efficiency. Another thought: though the above statistics date to 2006, it’s not as if the anti-malware burden has been growing lighter.
What I think might be unfolding is a divergence between Apple and Intel in computing-security philosophy. I don’t think Apple sees as much of a legitimate role for runtime security processes as Intel. As the leading exponent of the walled-garden model of computing (first for iDevices and, this week, for Macs), Apple seems to be fleshing out a vision where the average user’s kingdom is protected from malware entry at the point of purchase; meanwhile, downloads from elsewhere (at least for desktop and laptop systems) can be dealt with by scanning at the time of download and installation, as was quietly implemented in Snow Leopard in the summer of 2009 (http://www.zdnet.com/blog/security/apple-adds-malware-blocker-in-snow-leopard/4104). This leaves the processor to run unburdened by scanners and sniffers the rest of the time.
(Intriguingly, Microsoft has embraced the walled-garden model with Windows Phone 7.)
Personally, as a computing-security paranoiac, I like (my notion of) Apple’s vision better. Intel’s vision (assuming my analysis holds water at all!) requires them to think of everything ahead of time when they design their chips, unless they include some sort of updateable microcode or firmware or FPGA-like fabric, which would impose real-estate, speed, cost, and power penalties and might open the door to privacy concerns, considering the recent allegations of spooky back-doors in certain OSes (http://bsd.slashdot.org/story/10/12/15/004235/FBI-Alleged-To-Have-Backdoored-OpenBSDs-IPSEC-Stack?from=rss). And I’ve seen too many major antivirus engine upgrades necessitated by new tricks by malware authors to be comfortable with the notion of hard-coded, un-updateable runtime security technologies. Hard-coding works for GPUs because graphics and video standards are, well, standards, and developers stick to them. Malware authors are hardly so pliant. And then there’s the Stuxnet adventure, still unfolding, in which clever and very probably state-sponsored malware authors targeted specific machines using low-level vulnerabilities. Backdoor concerns will only harden after that.
Meanwhile, it’s pointless to build a back-door scanner into a lean, RISC architecture running a decent _nix-based user sandbox. Apple may be betting that it’s not necessary and, moreover, not advisable in a world where power consumption increasingly trumps all—which would mean a widening chasm between itself and Intel. That would mean more emphasis on Apple’s own chip technologies, portending the first Apple-branded processor in a MacOS (rather than iOS) machine in 2011.
Remember the first Apple machine of the Intel era was... the XServe. Which goes to the graveyard shortly. To be replaced by (rubs envelope on forehead)... the MacBlade, running MacOS in server trim, and sporting Apple-branded chips. Whether these will be ARM-based like the A4 or PowerPC-based and leveraging PA Semi’s know-how is the fun unknown to watch. But I’m betting the box won’t have that Intel Inside sticker. It will, however, be gaudily multicore, and will leverage MacOS’s Grand Central Dispatch in marvelous ways. A wildcard is virtualization, where Apple was late to the party but has a real opportunity to rewrite the rules by baking Type I virtualization into the fundamentals of OS X as it updates the Mach kernel for the new chips, whatever they turn out to be. But that might have to wait for OS XI. Gotta leave something for 2012 after all.
If something like a MacBlade does happen, then this architecture will trickle into the Mac desktop and laptop product lines over the next year or two, possibly starting with the next refresh of the MacBook Pro in mid summer.
A Tectonic Shift in Storage Technology: Where Oh Where to Keep All My Stuff, and Enabling Another Great Leap Forward for iOS
All our wondrous iDevices and their competitors exist because flash RAM has moderated in cost and mushroomed in capability over the past several years. But flash still has drawbacks, including limited write cycles, though some manufacturers (Apple among them) have proprietary tricks for managing flash lifetime to the point that it’s good-enough for storage in iPhones and iPads, and lately for hard-disk replacements in the svelte new MacBook Air. Where it’s not good-enough is for virtual memory swap-file usage. Consequently, iOS does not do swap files. Its addressable memory space is limited by the actual RAM installed. (I’m finding inconclusive information about whether swap files are enabled in the new MacBook Airs.)
The thing is, we’re already starting to see alternatives to flash RAM emerge—alternatives which may sidestep the write-cycle limitations of flash; for example, memristor technology. And, meanwhile, the first cell phone with phase-change memory has already hit the streets, from perennial Apple partner Samsung.
This indicates that fully-fledged virtual memory may come to mobile computing in 2011. An iOS that can address many gigabytes of virtual memory would kick applications up to the next level. (Originally I wrote, “an iOS or Android,” but, on reflection, struck out Android, because this is a case where Apple’s software/hardware integration would enable it to lead. Again.)
Side comment: I wouldn’t want to be a hard-disk manufacturer in 2011. Like an eroding seashore, only the higher reaches of the storage market remain accessible to those folks, and what we’re witnessing is a steep cliff that crumbles beneath the feet of players scrambling for altitude in a shrinking habitable area.
Macroeconomics May Impact Android’s Uptake
2011 is looking like it’ll be the year when a few macroeconomic bills come due. Liquidity will finally improve, and with it will come inflation fueled by wads of trillions of dollars of recently-printed funny-money acting as accumulated accelerant. If so, Apple’s economies of scale will help it maintain price competitiveness while already-struggling Android-based handset manufacturers will find their giveaway business models under immediate and severe pressure. All prices will eventually have to rise, but Apple will be able to resist the trend longest, putting real hurt on the Android world.
Put another way: Android’s ascension is predicated on deflated prices. But if there’s a stiff inflationary updraft, that might destabilize Android’s strategy, especially if Apple holds firm. So fast pump-priming might do the job Oracle’s lawyers are trying to do... and their efforts will heat up in 2011 too, further contributing to pressure on the Android “free” model.
A mid-2011 iPhone Nano would be very well timed, then. Aikido!
If correct, Android’s market share in the mobile market will peak in mid-2011, then stagnate or decline for at least a while as fewer hot deals are offered to consumers. A-to-B cost comparisons usually work to Apple’s advantage, after all. It’s one reason why there are no iPod competitors with any market share to speak of. Corollary: it will be doubly nice to be an iPhone carrier.
The FCC will be slapped down by courts (which already told them they couldn’t do what they just went and did) and abused by the new Congress, which is less wedded to Google campaign donations, less convinced that there’s any sort of problem that needs Federal fixing, and more inclined to regard Internet providers’ pipes as private property over which laissez faire principles should reign.
There may be things ISPs can do to still downrate traffic they want downrated—such as by providing (or requiring) routers or modems which implement traffic-shaping trickery that implements corporate prerogatives, but at the endpoints where Net Neutrality regulations would less apply. I’m old enough to remember how Ma Bell wouldn’t allow “unauthorized” fax machines and 9600-baud modems to be connected to its wiring... Plus ça change, plus c’est la même chose.
Apple’s Quiet War on Carriers Continues
Look for an official, built-in Apple VOIP application, perhaps a voice-only mode for FaceTime. It’ll come to the iPod Nano, too. I’m a little surprised that third-party FaceTime-compatible applications haven’t crossed my radar yet, given Apple’s promise to open the specification. Multi-user FaceTime should also emerge in 2011.
Sensors and User Experience Require Battery Breakthroughs
To me, the most dazzling aspect of computing’s incredible shrinking act is the sheer number and variety of sensors and gizmos being built into today’s handsets and tablets: gyroscopes, accelerometers, compasses, cameras (plural), ambient light sensors, proximity sensors, touch screens with invisibly small pixels, LED flashbulbs, touch screens, etc. (not to mention GSM, CDMA, WiFi and Bluetooth transceivers, and, soon, NFC). There’s a stampede to see who can jam the next technology from outer space into our pocket pals. Amazing stuff.
Up next: new battery technologies, with energy densities taking a manifold leap thanks to carbon nanotube electrodes. This will enable the integration of picoprojectors—perhaps not just to share images and videos with friends and customers, but also to project virtual keyboards—a trick I saw demonstrated nearly a decade ago, and whose day may be coming, as might eyeglass displays, which could bring 3D to your portable digital companion. Again, the pieces are juuuust-about-in-place as 2011 dawns.