Brian's Brain

ASUS was, I admit, one of the companies I was most looking forward to visiting last week (along with Acer, who'd been deleted from the schedule by the time I arrived in Taipei, and HTC, who canceled on us at the last minute).

The company has been notably aggressive at expanding beyond its ODM (original design manufacturer)-subcontractor roots and cultivating a retail brand name of its own, across numerous product categories…PDAs, smartphones, mobile and desktop computers, and motherboards and graphics cards for build-your-own PC enthusiasts. And its products not only target unique product and pricetag categories (i.e. the Eee PC series), suggesting a strong internal encouragement of innovation, they also exhibit eye-catching industrial design influences (aka Bamboo) that might even make Steve Jobs proud. As might the Mona-Lisa-Out-Of-Motherboards sculpture in the company's front lobby:

An upfront disclaimer; I'm pretty skeptical that Taiwan's companies are going to be able to navigate the transition from ODM powerhouses to retail brand names without substantial fiscal downside, at least in the short term. Take GIGABYTE, for example; company representatives admitted that ~80% of company revenues currently derive from the combination of PC motherboards and graphics cards, with much of this product shipped to other companies for retail brand and re-sell. I don't need to educate you in detail, I think, on the wholesale desktop-to-mobile computer transition that's already well underway worldwide. Consider the substantial impact that this evolution will have on both desktop PC-targeted motherboard and graphics card sales going forward, not to mention the impact of the recession currently gripping the United States and other key markets. And consider that, more generally, most of the product areas that Taiwanese suppliers are targeting for direct retail entry are already saturated with well-known brands.

With that said, I think that ASUS is comparatively well positioned for success in this changeover. As such, I'll spend the remainder of this writeup sharing with you the responses of company chairman Jonney Shih (whose candor and humour I found quite refreshing) to various questions I asked him during the post-presentation discussion.

• Regarding today's substantial Chinese manufacturing muscle and the likely eventual emergence of Chinese retail branding efforts in the future, Shih suggested that the ideal combination is 'Chinese cost plus Taiwan speed', and that he sees 'no big inherent threat from China'.
• ASUS has approximately 150 engineers devoted to IC designs for the company's various products. Although the overall corporate tendency is to employ other companies' semiconductor products, in order leverage volume cost efficiencies and other attractive factors, Shih notes that sometimes the company still finds that it can deliver a value-added end result by 'going it alone' from a semiconductor content standpoint.
• More than 500 engineers comprise the company's smartphone division, even though revenues in this particular product segment are currently 'small'. ASUS reportedly has key IP in the 3G-and-beyond technology arena, and Shih is a big believer in the near-future potential of MIDs (mobile Internet devices) and other supplements-and-replacements for the traditional laptop computer.
• On that note, I asked Shih for his impressions on the looming ARM-vs-Intel Atom fight for the silicon foundation of future smartphones and other portable communications-and-computing products. Shih feels that, in its initial 45 nm-fabricated iteration, Atom will do well in handheld computers but is too power-hungry for use in mobile phones. However, at 32 nm, 'the battle between x86 and ARM will truly happen'.
• Speaking of Intel, I queried Shih on his company's current impressions and future sourcing plans for Taiwanese neighbor Via Technologies. He suggested that, to date, Via has been unable to break free from the price and performance competitive shackles of much larger AMD and Intel at any particular point in time, although he appreciates Via's longstanding focus on cost and power consumption. To date, ASUS's use of Via has been restricted to 'in the lab'. Going forward, Isaiah (aka 'Nano') looks 'better', although he again is dubious that the company can sufficiently differentiate itself from the inevitable AMD-and-Intel competitive response.
• And what of ASUS's future AMD-based product plans? Shih was unwilling to state anything more definitive than offering an observation that 'Intel's future product roadmap looks very strong'. He also noted, as a case study example of the substantial resources that Intel is able to bear on a problem in order to preserve its market segment share dominance, that while the company's California- and Oregon-based design teams struggled for many years with the power-hungry and performance-strapped Pentium 4 product line, Intel's Israel-based architecture group was simultaneously developing the Pentium M and Core microarchitecture-based successors that now comprise the bulk of the company's products.
• Shih wouldn't give me any solid forecasts as to what degree the Eee PC's recent O/S broadening beyond Linux to Windows XP would boost product shipment volumes. He noted that the company put a lot of development focus into Linux, at least at first, and that 'Microsoft felt the threat' and responded by extending Windows XP's planned development and support life and by optimizing the O/S for small form factor systems' CPU, memory and battery capacity resources. Going forward, Shih sees Linux as still offering rapid market entry and feature set flexibility advantages to the company, and being applicable to 'early adopters and power users', whereas Windows provides 'familiarity' to potential customers.
• Shih's more optimistic than many analysts seem to be about the future of the desktop PC. He admits, especially now that the CRT-to-LCD transition is largely complete even in emerging markets (where he sees the bulk of motherboard business growth going forward), an all-in-one laptop is often a cheaper alternative to a two-piece PC-plus-display alternative system configuration. However, he points out that many potential usage environments value large displays that laptops can't support, and that the cost factor can also be mitigated by tethering an inconspicuously sized and located system to an already-present television.
• Regarding the future of the standalone graphics card in the face of the integrated core logic-with-graphics onslaught, he's also comparatively positive in his forecast. In spite of gaming-on-PC stagnancy, he points to expanded future use of the GPU in, for example, still image and video processing acceleration functions. And what of physics processing? When I pointed out that ASUS was a notable pioneer in offering a physics acceleration add-in card, Shih admitted that the company was an investor in AGEIA, thereby explaining at least part of the underlying motivation for this particular product experiment.
• Finally, in response to my query regarding Microsoft's longstanding Tablet PC concept advocacy (specifically) and various industry implementations of touchscreen interfaces (generally), he noted that some products in the company's line already employed touch capabilities. Whether touch would be the primary or supplemental interface going forward depended in no small part on the system form factor i.e. whether or not there was room for a functional keyboard (physical or virtual). As a general rule, ASUS would balance the incremental cost of touchscreen inclusion against the perceive added value of the feature in deciding whether or not to implement a touch interface in future products.