Disposable ATE?

IBM has announced that it is selling its PC business to China's Lenovo Group. Conventional wisdom is that IBM can't compete with the likes of Dell in a cost-sensitive market where supply-chain management matters more than technological innovation. The Wall Street Journal's December 6 "Real Time" column contends that IBM, a maker of "big iron," saw PCs as "little iron," instead of more realistically as "disposable boxes of cheap parts."

IBM's PC woes suggest possible parallels with ATE makers, many of whom offer personal-sized ATE systems (see "Tabletop testers: From prototype to production? " Test & Measurement World, September 2004) or at least modular, open systems that cost significantly less than their big-iron offerings. Will traditional big-iron companies falter in the marketing of lower-cost systems?

The evidence suggests not. First, low-cost ATE remains a far cry from "disposable boxes of cheap parts." ATE systems may have only a couple of years of life serving cutting-edge applications, but they don't quickly end up on the curbside. Seyed Paransun, VP of test business at Amkor Technology, told me Amkor is still using 20-year-old testers. Indeed, one application for LTX's tabletop Fusion DX is to augment legacy systems.

Second, IBM used off-the-shelf components and assemblies from outside vendors for its PCs (see "Whence came the IBM PC? " T&MW, September 15, 2001); the adoption of such components, especially the Intel processors and Microsoft operating system, opened the door for other computer makers to build compatible systems and compete on price.

That's not the case for ATE companies. Although they may outsource various aspects of their systems' production, they maintain tight control over tester architecture and operating system—even the relatively open Credence Sapphire and Teradyne Flex systems.

The possible exception is Advantest, whose embrace of the OpenStar standard opens the door for a competitor to launch a price war. Such a competitor has yet to emerge, however, and it's more likely that the OpenStar effort will attract instruments—such as Roos Instruments' OpenStar-compliant Cassini for RF device test—from multiple vendors, but not mainframes.

So, is there no room for the Dells of the ATE industry? Perhaps the closest analogues are firms like Inovys and Teseda, who have no "big iron" history or the baggage it might entail. They are likely to do well in the niches they serve (STIL-based DFT-focused digital test), but the traditional vendors will dominate for high-end mixed-signal applications.