Commentary: Will Transmeta gain traction in chip market?

Casio is the latest company with plans to introduce a sub-notebook based on Transmeta's Crusoe chip. Indeed, a number of Japanese manufacturers, such as NEC, Fujitsu and Sony, are developing Crusoe-based systems.

This week, Transmeta also gained a major infusion of

	See news story: Casio to adopt Transmeta for notebooks

capital from its hugely successful initial public offering, which defied the current unfavorable climate for new stock issues and raised $237 million. The stock was priced at $21 per share--higher than the initially expected offering price of $16 to $18--and closed Friday at nearly $41.

However, despite its short-term momentum, Transmeta faces major obstacles before it can hope to take a significant chunk of the microprocessor market away from Intel.

Foremost among these obstacles is Intel itself, which has already announced that it will introduce its own line of extremely low-power Pentium III chips for notebooks in mid-2001. We believe Intel will offer low-power chips that match Transmeta's power curve within 18 months.

We see this as a typical case of a start-up goading an industry giant into action. Transmeta has built its entry into the marketplace on meeting the real need for longer battery life in laptops. As Intel catches up (or even comes close) in the extremely low-power area, Transmeta will likely be forced to remain a niche provider of chips for specialized devices.

This would repeat the trajectory of other upstart chipmakers over the past decade, which after being greeted with hype and high expectations, have fizzled out when faced with the challenge of actually placing their chips in products that sell in significant quantities.

To address the formidable advantage Intel has in the application market, Transmeta provides x86 compatibility. However, despite Transmeta's claims that its low-power processor emulation technology provides the x86 compatibility needed to run standard Windows applications, any chip emulation system--such as Crusoe--requires extra central processing unit (CPU) cycles to run.

More importantly, Transmeta cannot guarantee that all Windows applications will run on Crusoe or that they will run at a given speed. Performance may vary from one application to the next. In upcoming months, Crusoe-based systems will undergo detailed benchmark testing by disinterested parties, which will produce a clearer picture of the degree of compatibility and performance achieved.

We believe the performance issues and incomplete application support make Transmeta-based systems a risky investment for buyers and vendors alike. But if Transmeta delivers truly unique value, the risks of using these chips will drop.

A key part of the Transmeta message and value proposition is its promise to dramatically extend battery life. However, early hype has yielded of late to more cautious speculation about the amount of battery life extension--pointing up the fact that battery life does not depend only on the chip, but on the entire system, including the disk drive, display and modem.

This reality has caused many of the mainstream PC vendors to either cancel or forgo development of systems using Crusoe. We see most investments in Crusoe coming from vendors using it as a differentiator as they try to break into a market, or from companies using it in emerging markets.

IBM's recent decision to suspend its plans to build a ThinkPad notebook based on the Crusoe chip was reportedly related to disappointment with the extension of battery life that, in fact, resulted. And Compaq Computer has reportedly decided not to use Crusoe in any of its domestic products, despite having a significant investment in Transmeta.

To maximize the benefits of its low-power usage, Transmeta chips are finding their initial applications in small devices such as sub-notebooks and computing appliances, where overall power consumption is relatively low, thus maximizing the potential battery-life benefit of a low-power chip.

This includes devices such as Via's just-announced "wearable computer" using a Crusoe chip or Gateway's new Web-surfing appliance called the Touch Pad, which will work only for subscribers to America Online. In parts of the pervasive computing appliance market, x86 emulation is less critical, increasing the viability of Crusoe.

Only in the unlikely event that Intel does not match Transmeta's chip performance in any significant way, would Transmeta garner meaningful market share in the mainstream mobile computing market. Otherwise, Transmeta will need to find a viable future as a niche provider of chips for small devices requiring low-power consumption.

We do not believe companies should be eager to support mainstream devices--for example, notebooks and handheld devices--using Transmeta's Crusoe chip or any other chip that is not native x86 compatible. If a system relies on processor emulation, it will not necessarily be successful at supporting a standard corporate application configuration, creating potential support headaches. As more Crusoe-based notebooks and computing devices reach the marketplace, companies keep a close on eye on who is buying them.

Meta Group analysts David Cearley, Peter Burris, Dale Kutnick, Val Sribar, Jack Gold, and William Zachmann contributed to this article.

Entire contents, Copyright © 2000 Meta Group, Inc. All rights reserved.