posted by Bryon Moyer
Consumer electronics is a tricky business. The scale is such that the rewards can be huge, but the logistics of manufacturing at that scale can limit who can jump in. Smartphones, however, have provided an opportunity for the Little Guy to leverage the platform by figuring out new and innovative ways to use the hardware that’s already there. The sensors in particular have been a rich source of innovation.
I’ve you’ve got a clever invention, then, assuming it’s all software, all you have to do is make sure it’s rock-solid (hopefully, if that’s convenient and you feel like it) and then put it up for sale, where it will compete with only hundreds of thousands of other apps for your target customer’s attention. In this case, access is easier, but standing out is hard.
I recently saw an invention announcement that falls somewhere in between. It has the feel of a small guy trying to impact the actual hardware of a cell phone. The company is Fraden and the application is the ability to take the temperature of a person or thing simply by “taking a picture” of the item.
The classic “Awwwww” picture that they show is that of a camera taking a baby’s temperature. (Although it’s clearly a chopped photo…)
But the announcement wasn’t about a new application that you could buy, nor was it about a new phone platform. It was about technology ready for license – and it likely involves adding hardware to the phone: an infrared sensor. The rest works with the camera in the phone. If the camera auto-focus were active, using an infrared ranging system, then all the hardware would already be in place (assuming the software had access to the sensor). But most cell phones use passive auto-focus, so this means adding a new piece of hardware to the phone (albeit one claimed to be inexpensive).
That’s a much harder row to hoe. You can’t go directly to the consumer: you have to convince phone makers to add yet another sensor to the laundry list they already have. How do you do that? By trying to tap into phone-makers’ desperation for differentiation. The phone that can accurately take a kid’s temperature is likely to have an advantage, at least in the people-with-kids demographic. (To the extent that anyone gives a crap about people over 20 as a useful target market for gadgetry…)
Of course, as the inventor, you really want to show up in all phones. Which actually means modulating the differentiation message as penetration changes. At first, you can say, “This is huge for differentiation: you’ll be the only phone that has one!” After a couple successes, that changes to, “More and more phones are differentiating themselves with this feature.” Until finally the message morphs to, “Everyone else has this: you’ll be the only one without.”
In other words, differentiation enablement as a selling point can only work at the beginning of the campaign. If you’re completely and utterly successful in penetrating the market, then, by definition, your differentiating technology has eliminated any differentiation.
You can read more about the Fraden story in their announcement.
posted by Bryon Moyer
There are two widespread myths in the MEMS world. Or so said ST’s Benedetto Vigna, EVP and GM of ST Micro’s Analog, MEMS, and Sensor group, at the recent MEMS Executive Congress.
The first, the primary topic of this note, is that MEMS needs standards. In fact, there has been a hue and cry for MEMS standards for a while now, although there’s less clarity on exactly what needs to be standardized. Discussions are ongoing (and we’ll look more deeply at this in an upcoming article), but it would seem that, while that happens, ST is pretty convinced that things are just fine as they are.
This gets to the central question of who benefits from standards. Ideally, when done right, the customers benefit, the market grows or becomes more efficient, and there is more business to chase for all players. But sometimes, for a given company, it may feel that standards will simply benefit the competition.
While not stated in those terms, it could be inferred that ST, being a dominant IDM in the MEMS world, sees standards doing more of the latter than the former. As Benedetto explains it, the only real standard needed in order to make users’ lives easier is the API. And that happens at the OS level. And while the sensor details burble along far below that level, enough low-level stackware should be built to allow users to work at the OS level. Nothing else matters.
This means, of course, that he sees ST capable of meeting all of these customer needs on their own. Certainly as a large company with established success and control over their own production, they’ve got (or can make available) the resources to do this, and presumably they view that as a competitive advantage. And it probably is.
Unstated, but perhaps implied (or at least easily inferred), is the position, “We’ll simply make everything the customer needs, they’ll buy everything from us, and no one else will matter. So standards will be irrelevant.” They’re filling out what he calls the “five fingers” of the MEMS UI: accelerometers, gyroscopes, magnetometers, pressure sensors, and microphones. From this standpoint, if they’ve got all this covered, then it’s a done deal.
Separately, he noted the second myth – and it could be said to be related: Improved EDA tools are important for designing the transducers. While we won’t dig into this one here, it could also be perceived as a natural position to take when you’re a big guy with a long history and lots of internal processes, tools, and experience under your belt. Better EDA tools then only make things easier for the competition.
It’s always a tough timing call when you’re a big player with an early lead. For a while, you can operate like you’re alone in the market. In fact, in the early days, you more or less are. During that phase, you can resist calls for you to do things to make life easier for the newcomers. In the perfect world, those newcomers would dissolve away and you’d be back in charge of the market.
That rarely happens, of course, so there’s that delicate decision to make as to when to move into a more conciliatory position. From a practical standpoint, that’s when you shift from ignoring or even blocking the standards efforts to working on them to minimize damage.
At this point, of course, I’m completely putting words into their mouths – they’ve said nothing about these last points specifically. I’m pulling this from my own experience with standards many years ago. But it will be interesting to see how they and the other big MEMS IDMs jockey around as standards efforts coalesce.
posted by Bryon Moyer
Karen has been feeling unloved.
That would be Karen Lightman, managing director of the MEMS Industry Group. After several years of MEMS taking a share of the attention at Semicon West, she noted a distinct “There’s a new kid in town” feeling this year in her blog, and this turned out to be a very well-attended blog post.
So as a follow-up, she arranged a webcast for last week that featured Peter Himes of Silex, a MEMS foundry, Mike Rosa of Applied Materials, and yours truly. While my role was more to provide an outside view of why we at Techfocus Media like MEMS (and noting that such gushing is actually pretty unusual for me since I often feel more at home as the curmudgeon in the corner), Peter and Mike focused more on the internal details of the industry.
The question gets to the relationship between the MEMS industry and the IC industry. It’s been noted that the downturn might have boosted MEMS by giving fabs another way to fill their lines while chip sales were slow. The question is, when the economy ramps back up, will that put MEMS back out in the cold? Are IC makers no longer interested in MEMS? (This is, of course, a highly simplified synopsis of the question.)
The thing that took the spotlight at Semicon West was 450-mm (roughly 18”) wafers. And wafers (at least unpatterned ones – and no one has been patterning 18” wafers yet) are shiny. And everyone loves something new and shiny. But MEMS doesn’t need – couldn’t come close to justifying – 18” wafers. Some say the physics won’t even permit it (these are mechanical systems, after all – stresses across such a huge wafer matter).
While the question was somewhat limited in scope, my take on it was that it went to the whole existential justification of MEMS. After all, if a key part of the supply chain – the foundry – goes away, well, you don’t have MEMS devices. So while the question wasn’t, “will MEMS go away?” it seemed to me that, following the thread of logic, that’s where you end up. Which is why I focused my energy on the fact that I don’t think MEMS is going anywhere. People expect the things that are possible due only to MEMS, and this could not go away quietly. And we all agreed on that.
All of these issues were addressed during the webcast, but there was one other thought that came to mind that we didn’t get to. And that’s the question of whether everyone is going to run headlong to 18” wafers, which would also leave MEMS looking for a dance partner. But these wafers are only for super-high-volume super-low-cost devices – and even having said that, the future of 450-mm blanks is not foreordained.
When I discussed the topic with KLA-Tencor some months back, I got the sense that a) everyone was excited about jumping into the 450-mm pool, but b) everyone was standing on the edge of that pool waiting for someone else to jump in first. And when KLA-Tencor jumped in with their blank wafer inspection tool (the first necessary step before going any further in the production chain), they even hedged their bets by providing two docks: one for 300-mm wafers and one for 450-mm wafers. That way a company wouldn’t have to justify the purchase based solely on 450-mm wafer demand.
The number of people that had signed up for the webcast indicates, however, that these questions caught people’s attention. So hopefully, minds have been eased following the discussion.
Karen certainly seemed happier.
You can watch a rerun of the webinar here. (With apologies for my being over-optimistic about the current ability of the internet to handle PowerPoint animations...)