When is a teardown not really a teardown?

Do you need to tear it apart to know what's inside? Not always according to one analyst firm.

For many Apple lovers and haters alike, the next product launch and nearly coincident product teardown is always an eagerly anticipated event. As the heat continues to give gadget hounds mid-summer dreams, there are expectations of a new iPhone, iPad, even a complete panel TV adding the display portion to the existing “hobby” line of set-top boxes. But the dreamers have other manufacturers to rely on for product launches to watch, teardowns to pore over, flames to post, and perhaps even the odd meaningful discussion.

Perhaps it was the natural symbiosis of the media and Apple Inc. marketing, but it seems to me that the volume of press coverage of product teardowns has more or less tracked the rise of that company over the last half dozen years. But there is more than just the unveiling of products to see what makes them tick. We usually have some early knowledge of the primary components like micro / applications processors (APU, CPU whatever the flavor of the day), but we thirst for more. Why? For hackers and the repair-minded, there will always be some desire to understand how to “get in there.” Since semiconductors and integrated circuits are my own primary interest, I appreciate the inherent value of seeing the winning manufacturers for all the sockets designed into a product. Perhaps I am not on the margins after all. Who else is interested?

iFixit Teardown Image
                                                                                Image courtesy iFixit.com

For the more detailed list of included components, several information purveyors offer spreadsheets of estimated component costs along with their product teardowns that attempt to calculate the total cost of materials for the product. These are known as BOM (bill of materials) reports.

Nowadays, the headline for a product teardown is inevitably the price of the components, but these BOM calculations find their way through media as a convenient aid to marketing and PR for the firms who perform them to raise their profile. (In one case, the media outlet is owned by the same entity as the teardown analysis branch.) For average users, hobbyists, aficianados, even hackers, the total estimated price of components is not that meaningful – unless you get a kick out of seeing how much you are personally contributing to the product company’s bottom line. It’s the kind of information that excites accountants, not engineers.

Whether it was the rush to grab the first headline or just the repetition of looking at the same basic design again and again is hard to say, but UBM TechInsights announced a $184 BOM estimate for Google’s first tablet – the Nexus 7 – before even getting its hands on one. The newsworthiness of this announcement lay in the fact that the Google tablet was significantly higher cost than the Wall Street Journal reported as Amazon’s “identically priced Kindle Fire.” The analysis leading to the extra $30 or so was based on a higher resolution display and the choice of NVidia Tegra 3 processor rather than the TI OMAP 4430 found by iFixit in the Kindle tablet. Come to think of it, everybody knows iFixit will have the teardown first. For everyone else, they need to find other ways to appear to squeeze into the news cycle.

The UBM TechInsights un-teardown drove home something I had been thinking about with respect to Apple and its competitors. With apologies to Research in Motion, the only remaining tablet or smartphone hardware company with a proprietary operating system is Apple. Even Nokia is moving to Windows and everyone else relies on Google for Android. The hardware for these “freely” available software platforms tends toward the generic too. I would liken the hardware platform design of Android-based tablets and smartphones a standard approach not unlike the old beige box PC clones. You just don’t expect to see anything significantly different from one to another. And it’s not particularly surprising considering how these manufacturers rely on outsourced partners for key pieces (hardware in Google’s case, operating system for the traditional hardware companies).

I fully expect this to add another bar to my meritorious Apple fan boy service medal, but the relative ease of estimating the price of components of the Google tablet without even cracking it open is food for thought in comparing alternative product design approaches.

One of the debates that BOM estimates always open when analysts attempt to compare prices paid for components by one manufacturer compared to another is the volume buying and concomitant price pressure that the Amazons, Apples and Googles can bring to the table. In differentiating the “open source” products from Apple, there are three key features to any Android (and to a lesser extent Windows-based) product:

1. Volume component buying pressure applied to vendors
2. An “open source” operating system with some individualistic flair provided by the manufacturer
3. The industrial design of the handset or tablet computer.

In Apple’s case, these are taken further and augmented by additional strategic pieces. As for the OS, Apple goes much further with a completely proprietary software stack that is supported by a large and vibrant developer community producing third party applications. But there is more:

4. Complete in-house design and control of the hardware system platform
5. Integrated circuit chip design for key components.

This last piece is intentionally shown in the last spot since it is the last piece of the puzzle rolled out by Apple. If you are looking for a quick flame for this post, you will notice that the fifth item describes “components” plural. Yes, Apple has so far introduced semiconductor designs for only the applications processor. That’s only one. Singular. But will it be the last? That seems doubtful. More likely, you should expect to see the most change in Apple products coming out of the semiconductor design group since it is still early days. Apple introduced its own branded A4 applications processor chip in April 2010 with the launch of the iPad. Hard to believe but the iPad has only been available for a little over two years.

Control of the most critical piece of the hardware allows Apple a lot of flexibility in optimizing both the hardware and the software for the consumer application instead of depending on the “one size fits all” approach that is unavoidable for vendors hoping to win the most platform sockets amongst the widest range of consumer products.

Apple’s semiconductor design team is captive. Their goals are singular and forced by corporate management to synchronize with the operating system, user interface, and platform and industrial design teams. Apple haters will say that anything “closed” is bad. Investors and users alike are free to disagree.