Commentary: At a recent chip trade show, Transmeta chief executive David Ditzel was yapping about how he made Intel boost its mobile R&D budget. Meanwhile, he carried around plenty of models featuring Transmeta chips.
Transmeta's goal? To be a thorn in Intel's side as it conquers the world of mobile computing chips. For now, Transmeta is just an annoyance. Nevertheless, Transmeta's moxie is getting its IPO lots of attention.
The company is planning to launch its 13-million-share initial public offering Tuesday with a price range of $16 to $18. The range was bumped up Friday from $11 to $13. Morgan Stanley is the lead underwriter. Transmeta's software-based Crusoe chip has garnered plenty of hype in the tech world. The fabless chip maker's pitch is that it can make semiconductors that run cool and conserve energy without sacrificing performance.
To be sure, Transmeta has a good story. But it's a story that isn't past the second chapter. Wall Street is clamouring to get Transmeta shares, and that guarantees a big opener. But the individual investor, who isn't going to get the IPO price, has to ask the tough questions before going ga-ga over Transmeta, which still has a lot to prove.
For the six months ended 30 June, Transmeta reported sales of $358,000 and a net loss of $43.7m. For the quarter ending Sept. 30, Transmeta said its revenue would be about $3.5m. The company is lugging around an accumulated deficit of nearly $120m.
Transmeta's product history can be described in one word -- brief. In January, Transmeta unveiled its Crusoe chip. By the end of the first half of 2000, the company made limited quantities of its chip. In September, Transmeta shipped in volume. IBM manufactures Transmeta's chips.
Now the fun really begins. All Transmeta has to do is convince notebook computer manufacturers to use its chip. Step aside Intel and Advanced Micro Devices. In regulatory filings, Transmeta noted that "substantial resources" will be devoted to educating potential customers. If the company can't convince manufacturers to use Crusoe -- and potentially redesign products for Crusoe -- Transmeta is cooked. Although Transmeta has some big backers -- AOL and Compaq -- nothing is guaranteed.
Last week, IBM canceled a project aimed at building a ThinkPad mini-notebook using Transmeta's Crusoe processor.
Since tackling the notebook market is a tall order, Transmeta is also planning to be the chip behind a host of Internet appliances. The catch? The wireless and wireline technologies that would make appliances commonplace are still being developed.
Basically, if you buy into Transmeta's IPO, you're buying into a hunch. "There is little historical financial information that is useful in evaluating our business, prospects and future operating results," the company said in filings.
A few points to consider about Transmeta's story:
- Competition: Intel and AMD are bad enough, but those foes just cover the notebook market. On the Internet appliance side of the equation, Transmeta will battle MIPS, ARM Holdings and National Semiconductor.
- Breaking into the market: Transmeta may have a better mousetrap, but its competitors are entrenched and some aren't going to budge. For example, Dell is so loyal to Intel that it won't offer AMD chips despite the fact all of its rivals have AMD-based products. Transmeta can educate customers about its Code Morphing software and the benefits of its chips, but standards aren't built overnight. To its credit, Transmeta has a "working relationship" with Microsoft and consults with the Linux crowd to ensure compatibility.
- Small customer base: As of 30 September, Transmeta has shipped products in volume only to Sony Electronics and Fujitsu Ltd. Sony will use the Crusoe chip in its VAIO PictureBook C1VN notebook computer, and Fujitsu will use the chip in two of its notebook computers.
Hitachi is signed on to use Crusoe in three notebook computers and an Internet appliance. Meanwhile, Gateway will use Crusoe for Internet appliances still under development with America Online. It's a good start, but Transmeta needs to land more customers.
Despite Transmeta's undeveloped financial picture, many betting folks will bank on chief executive Ditzel, who has one helluva resume. He co-developed the reduced instruction set computer, or RISC, microprocessor technology while employed at AT&T's Bell Laboratories and was director of SPARC Laboratories and chief technical officer of Sun Microsystems' Microelectronics division.
Ditzel also brought a few brainiacs along for the ride. Through September, Transmeta had 109 employees with postgraduate degrees, including 41 PhDs, out of 364 employees.
See ZDII for US tech investor news.
See techTrader for more technology investment news, plus quotes and research.
To have your say online click on the TalkBack button and go to the ZDNet News forum.