Thursday, January 8, 2009

On the Palm Pre

The Palm Pre introduction at CES '09 today both impressed and informed my understanding of the handset industry.

Palm has put together an impressive, differentiated product. 

They use the latest OMAP CPUs from TI, model number 3430. The chip is part of TIs "High Performance" segment. The OMAP CPU is really a "System on a Chip" (SoC.) It has multiple cores. The OMAP looks like this:

Apple's iPhone chip is harder to find info about, but it is best understood to be the ARM 1176JZ(F)-S. Look here for more on that. Both chips feature the ARM CPUs. The 3430 has the newest "Cortex" CPU, which is a newer iteration of the ARM11 family CPU core featured in the iPhone. The 3430 is a 600 Mhz part, though Apple and other hardware makers seem to clock the chips down in practice. Nokia also uses the OMAP SoC line, so I assume they will get a crack at the 3430 soon. Nokia outsells Palm by multiple digits after all. The OMAP is certainly competitive.

I am less able to assemble a report on the other cores. The vagaries of OpenWhatever X.X this, "streams", and the difference between OS X and Palm WebOS GPU implementations means it comes down to gut. Apple has more experience here, and has proven games shipping that are competing with the PSP graphically. I give Palm second place, but a knowledgeable reader might know and have real facts to back it up.

Palm has built WebOS to compete in the software realm, and it is impressive as well. 

The aim is clearly to give users the best of all worlds. Desktop-style application architecture, broad interoperability with other OSes and relevant web services, and a slick, non-ideological interface. 

Palm Cards is Spaces done right for iPhones. No spatiality, rather, something like an "all windows Expose" of multiple concurrent Finder sessions, organized in a Vista-ish stack view. (Palm doesn't moronically overlap them, though. Microsoft renders the metaphor literally and makes the second page mostly covered by the first, as it would be in a stack.)    On a PD-, ahem, smartphone, the sessions are simple enough to be nearly iconic even at a small size. That is not true for desktops, and I am not suggesting this beats "desktop Spaces", but it does beat the almost Classical Mac dogma of single-tasking, pre MultiFinder (prescient name) reigning over the iPhone.

WebOS applications are akin to pre-SDK iPhone web apps, but more liberal with the hardware. Or, that was my interpretation. Palm may introduce signed apps or something like the app store and shift directions. Palm is small enough that a WinMobile-alike model with independent software houses building businesses on shipping free or paid applications makes more sense. It is also the model that Palm has used in the past. Newton, too, really. Perhaps Sprint will muscle a share of the overall application revenue stream, something AT&T would dearly like to have done. Apple gets a spiff from AT&T for signing up iPhone customers, and then Apple gets all the money for the programs shipped across the network. Surely Sprint will not allow wireless app delivery without exacting some tariff.

So the software is good, at least in demo form, and the hardware appears performance-competitive. The industrial design is solid, but relatively uninteresting, besides the cool inductive charger, and make-or-break (for some) hardware keyboard. Hopefully it is reliable. I have little doubt it will be a solid piece of hardware, at the least competent, and perhaps excellent, given Palm's culture and history. 

The Palm V was sublime. It is next to the iPod Nano in my mind as most transcendent, paradigm-forming iteration of handheld technology. Palm Pilots were impressive hardware, sized for the hand, unlike the relatively-enormous Newton. The Palm V seemed an impossible feat, it surely couldn't exist given its features and size. In the hand, it was perfectly sized for use, and felt extremely solid. The design was visually appealing, with a well-executed metal finish. 

[The iPod and iPod Mini were great MP3 players, the latter a logical extension of the flagship player. The Nano was a "whoa they can fit the iPod Photo into the battery of the iPod Mini" moment, and doomed hard disk MP3 players forever. The iPod was already second place in sales behind the Mini, though mostly because the iPod was still prohibitively expensive, and the Mini offered cheaper entry to the already-dominant iPod/iTunes ecosystem. The nano made the disk iPods (and especially the Rokr) seem quaint, solving the problem of carrying lots of music, sure, but not addressing how best to listen to music. From then on, the iPod was classic(al)...]

Jon Rubenstein is also on the scene, and he seems to have been given enough leeway to shape Palm up. The Treo x00s of the last few years were stagnant. The industrial design was cutesy 90s blobby pablum, and Palm screens were sunken and small. The Treo Pro demonstrated that Rubenstein could make the company move, advancing the form more in one shot than Palm had in many generations. The software, though not Rubenstein's avocation at Apple, has also made a Radical leap forward that is not simply a coincidence. If Palm dies, it will not be the former Apple executive's fault. The Pre is the result of real management chops.

Hardware, software, OS, all a go. But what about price? Palm seems unable to get it below the cost of an iPhone, and their network partner is not advantageous, let alone better enough to justify a large price premium. I bet the Palm hardware in a tear-down would be a bit over the cost of the iPhone, but not by enough to explain the retail price disadvantage.  

People forget, though, that Apple orders parts by the million, and has a specifically advantageous position when buying flash memory. Apple has every advantage in this market. Moreover, their OS development is funded by app sales, and addresses a far broader range of markets than does Palm's WebOS hardware lineup of one. The XServe and iPhone are both improved when Apple iterates Mac OS X. Expecting the Pre to equal the iPhone is like asking for a PowerMac at the price of a Dell in 1998.

The Pre will do a lot to make Palm competitive again, which is to say make it attractive to future large-scale investors like their current main squeeze, Elevation. I also think it will get Sprint some buzz, and buzz that resonates and snowballs. The G1 and, to a lesser extent, the RIM Storm, got buzz, but only a burst of noise that withered with real-world usage reports. The Pre could be good enough to get people talking about switching cell carriers for it. Friends will tell friends, the product experience will be different, good, and rare/low-volume enough to give the Pre some cachet.

These competitors are the ones that Apple should fear. They don't need to "knife the baby" Palm, but it is a left-field, well done, differentiated product that will make the iPhone shine begin to fade. Nothing is better now, but once the Pre ships mobile users will have a choice that is damn close.

1 comment:

jellinek said...

Great note. I think the key, as you point out, is price. Much like the old PowerMac, it has to be more expensive than an iPhone and thus won't capture more market. But Palm has impressively Thought Different, and might be better (at least v. the current iphone - who knows when it ships...).