Architecture Exploration of Processors and SoC to trade off power and performance 5
WP_Term Object
(
    [term_id] => 18
    [name] => Intel
    [slug] => intel
    [term_group] => 0
    [term_taxonomy_id] => 18
    [taxonomy] => category
    [description] => 
    [parent] => 158
    [count] => 365
    [filter] => raw
    [cat_ID] => 18
    [category_count] => 365
    [category_description] => 
    [cat_name] => Intel
    [category_nicename] => intel
    [category_parent] => 158
)

Burned again – can smartwatches recover?

Burned again – can smartwatches recover?
by Don Dingee on 08-05-2016 at 4:00 pm

With Intel’s Basis Peak smartwatch on intergalactic recall and Apple’s smartwatch sales falling faster than Skylab, designers now must face the real questions: 1) what is the right use case, and 2) what are the right chips to implement it?

Our Daniel Payne pointed out recently that Intel got blown out of the digital watch market once – in 1975, when TI put on a full-court press and designed a single chip watch for under $40. Analog watches were built into consumer culture, powered by years of John Cameron Swayze pitching Timex with “it takes a licking and keeps on ticking.” The leap to LED displays was easy, they were just fun to look at during the day and got better at night. Admit it – if you had an LED calculator, you typed in 7734, handed it to a friend upside down, and laughed like Beavis. (OK, maybe that was just me.)

Smartwatches are a much bigger leap of faith in utility. Before Apple hit the market, Pebble had the first big success. Still, the category was not quite defined yet. I saw the Basis B1 at CES 2013, thinking it was pretty cool. There was a whole section on “digital health” at that CES, with all kinds of fitness trackers. The formula seemed to be tethering the smartwatch to a phone app and including sensors to track personal health. A looming factor was NFC and the possibility of digital payments, access control, ticketing, and other uses.

Three years later, and what was supposed to be a hockey stick has turned into a niche-trough. The most recent smartwatch vendor roster shows how the mighty have fallen and the rest are CAGRing from relatively small bases:


I remember getting my chops busted on Twitter when Apple released its second version of watchOS, fixing some major flaws. My suggestion they had “missed” went over like a lead balloon. (“Just wait for watchOS 3, you’ll see.”) And yet, numbers never lie. Apple has failed to attract a sustainable user base, and there is no revolving upgrade door with next year’s faster, bigger, longer lasting model. Early adopters are in, mainstream users are not.

Fitness users, represented by Garmin on this list, still seem to have some interest for more rugged, multifunction devices. There’s also the elder care niche, well represented with the UnaliWear Kanega. The right niche can be profitable at the right scale – but it’s not what Apple, Samsung, and the big mobile guys are interested in. When Apple jumped in the pool, the splash knocked Pebble around, but didn’t drown them.

So, I had to laugh when I saw this chart in the Linley Group Mobile Conference introduction. I confess, I wasn’t there to hear the commentary that went with this slide. 38% CAGR looks like a Hanson Brothers autograph hockey stick to me:


There’s only one path for Apple to get it right and make that first assumption true with an upcoming release: get out of luxury mode and chop the price. The features/benefits equation simply does not add up at $299 and higher, even when watchOS 3 cuts in. At $50 as Linley did suggest in previous slides, things would be much more exciting.

Therein lies the problem. To get the price down, people are grabbing off-the-shelf microcontrollers, such as the Silicon Labs EFM32 Wonder Gecko with an ARM Cortex-M4 core in the Basis Peak. No MMU in sight, Android Wear and watchOS is still Cortex-A territory. Maybe the CPU is low power, but Basis managed to create a thermal problem somewhere – I’d guess in the battery charging subsystem, that’s speculation at this point.

No, we still do not have a chip co-optimized for a smartwatch operating system and priced at something like $10. Apple, Samsung and others are trying to add features to hit a cross-section of users who become enamored with something to justify their bigger chip price. The MediaTek MT2601 and Ingenic JZ4775 (based on a MIPS core) are finding some smartwatch sockets for Android Wear at lower prices. (Is there a RISC-V smartwatch chip in the making somewhere?)

I don’t see Apple adopting the TI mass market pricing strategy, it goes against their culture. Apple will probably recover from this trough, but won’t drive the 38% for a prolonged period. Ditto for Samsung Tizen. Intel should have paid a lot more attention. That leaves the fate of the smartwatch up to RTOS implementations like Garmin and Pebble, and Android Wear, and optimized chips we’re still waiting for.

Share this post via:

Comments

0 Replies to “Burned again – can smartwatches recover?”

You must register or log in to view/post comments.