The company introduced its low-power Crusoe processor earlier this year, after years of complete silence had led to rabid curiosity about its designs. And the company has managed to secure several key design wins in ultralight notebooks, despite a reluctance to publicly disclose the Crusoe's performance benchmarks.

Transmeta shares closed the day on Tuesday at $45.25, gaining more than 115% from the stock's initial trading price of $21. In early trading today, Transmeta shares rose higher to $46.36. Prior to the IPO launch, Transmeta's initial trading price had swelled, as the first price window for the offering was listed at $11-to-$13 one week before they began to trade. That figure was optimistically bumped up to the $15 to $16 range on Friday, and then bumped again to $21 the day before they began to trade on the Nasdaq exchange. By selling 13 million shares, the company garnered $273 million for its own coffers.

The enthusiasm of investors was not dampened by the recent decision from IBM not to use Transmetachips in its ultralight ThinkPad systems. IBM will continue to use Intel Corp. chips and also remains the sole manufacturing source of the Crusoe chips. "IBM was totally through with the design cycle and they were ready to start making those systems," noted Nathan Brookwood, principal analyst for market research firm Insight64 in Saratoga, Calif. "It has become clear that the Transmeta design sacrifices computational performance in order to get its low power consumption. It is unclear whether the market will accept this trade-off."

At least for now, it seems the stock market is more than accepting of Transmeta's vision, but some observers remain skeptical of the company's long-term prospects. To date, the company has been alone in the low-power niche it has defined as its prime target. But the 800-pound gorilla of the microprocessor segment smells some tasty bananas in the low-power segment, and Intel Corp. has already announced plans to develop its own line of low-power processors for the notebook market. Analysts suggest that upstart Transmeta will have its hands full competing with a company that has a history of steamrolling competition.

"We have the only microprocessor of its kind," said Bryan Hurst, director of worldwide sales for Transmeta, describing the Crusoe chips and their combination of hardware and software that delivers low-power chips compatible with x86 architectures. Although not a true x86 product, the Crusoe uses a unique code-morphing feature that translates x86 instructions into a format the chip can process, and runs on less than 1 watt.

In contrast, Intel's current line of Pentium III processors designed for the notebook market runs, on average, on about 2 watts, while a special, low-power version runs on 1 watt. Its low-power Celeron device for the mobile market also runs, on average, on about 2 watts, according to Intel data sheets.

However, Intel has disclosed plans to develop a new processor architecture, designed specifically to offer low power consumption, which could be available in the 2002 or 2003 timeframe. "We think the mobile PC market is growing enough that it makes sense to offer different products designed just for that market," said Frank Spindler, vice president of the Intel architecture group and general manager for the mobile platforms group.

Spindler explained that the company's 0.18- and 0.13-micron processes are a key part of its plans to reduce power consumption and that they expect to launch a Pentium III chip running at speeds over 1 GHz, with power dissipation of about 1.5 W by mid 2001. Pushing into 2002, he expects to see 500- and 600-MHz processors that dissipate less than a half-watt. The company will begin shifting to the 0.13-micron node in the first half of next year, which will further reduce power consumption.

All of this activity may likely threaten Transmeta's place as the sole vendor of ultralow-power chips. "Intel is certainly getting concerned about Transmeta," noted Linley Gwennap, principal analyst for market research firm The Linley Group in Mountain View, Calif. "Intel is starting to realize that there is a market demand out there."

Gwennap noted that Intel has not always been the first company to recognize a market opportunity, but once it decides to shift strategy, it is aggressive about moving into and trying to dominate a market segment. This is exactly what happened a few years ago, as Intel rival Advanced Micro Devices Inc. began to make a play for the low-end systems with its line of K6 chips. AMD was followed into that space by Rise Technology Co., Integrated Device Technology Inc. and National Semiconductors Corp.'s Cyrix unit.

Intel finally made a play for that segment by introducing the Celeron chip in 1998, and after a slow start, Celeron began to gain significant market share. That year was marked by price wars in the processor market and significant financial losses by AMD. After failing to ever win a major design win, Rise has quietly shifted its strategy to the embedded processor market. IDT and National have both sold off their processor divisions, and AMD is seeing success only after introducing a high-end processor, the Athlon. "Intel may be slow to move, but it usually hits its target," Gwennap said.

While Rise and IDT never really seemed to take off, analyst Brookwood points out that Transmeta already has some solid momentum. Besides design wins at four notebook vendors, the Crusoe is on deck to appear in a system produced by Gateway Inc. in a joint venture with America Online. "Transmeta has lined up some reasonable OEM partners," he said. "It clearly has a stake in the ground."

Transmeta's Hurst notes that while Intel may be detailing its road map for low-power devices but said there is a big difference between showing off a road map and actually delivering a product.

However, Gwennap said that Intel's history demonstrates that it would not be showing a road map if it had no intention to deliver those products. "Today, OEMs are flocking to the Crusoe because there is no real alternative at those power levels," he said. "But once Intel offers a competitive product it will become a more complicated issue, and Transmeta will have trouble holding onto the business it's got now."