October 2, 2001 9:20 PM PDT
Transmeta goes after non-PC chip market
"By this time next year, it could equal the notebook market," Mark Allen, Transmeta's CEO, said of the company's prospects in the market for embedded chips. "They are firming up product plans. They are still going through the evaluation process and benchmarking."
Embedded computers are systems that are not PCs, including set-top boxes and point-of-sale devices such as cash registers. The push into the embedded processor market comes as a way for Transmeta to tackle one of its chief longer-term challenges: volume.
Although the company landed a number of high-profile deals with major notebook manufacturers in Crusoe's first year of availability, analysts say Transmeta runs the risk of being wedged into a niche market.
The company sold approximately 500,000 chips in its first year of delivering products, according to Dean McCarron, principal analyst at Mercury Research. While that's an impressive debut, companies eventually need to hit about 1 million units in annual volume just to stay in the market, he said.
"One of the benefits about the embedded market is that the contracts last a long time. It can be a very lucrative market," McCarron said. "You can be partly successful in two markets and be a successful company...I'm sure a big part of the consideration is getting the volumes."
Transmeta will outline its new product road map later this month at the Microprocessor Forum. Right now, the company sells a 3000 line of Crusoe chips for Internet appliances and a 5600 chip for notebooks. In the fourth quarter, these lines will be replaced by the Crusoe 5800, a higher-performance version of the 5600 that also will cost substantially less, Allen said.
Next year, the Santa Clara, Calif.-based company will come out with an inexpensive system-on-a-chip that fuses a processor, a chipset and a graphics chip as well as a new high-performance version of Crusoe, said Dave Ditzel, Transmeta's chief technology officer.
The less-expensive integrated chip will take up about one-third of the space of the combined chips used for these functions. That chip will be used in servers, Internet appliances and embedded equipment, Ditzel said. The performance chip, meanwhile, will head into high-end notebooks. Samples of both chips will appear soon.
"If you can eliminate two or three chips out of the system, the savings can be substantial," he said.
All of the activity coincides with the first anniversary of Crusoe's arrival on the market. Although the company unveiled the technological theory behind Crusoe in January 2000, products containing the chip didn't hit the market until Sony released a Vaio notebook in Japan in October 2000. Most analysts hadn't even handled a demonstration unit until after the company's IPO a few weeks after the first chips came out.
Off to a good start
In many ways, Crusoe has had a fairly strong maiden year. Nearly every major Japanese notebook manufacturer, including Sony, Toshiba and NEC, has incorporated the chip into its domestic notebook lines. Although Gateway remains the only major U.S. company to incorporate the chip into a product, more deals are inevitable, Allen said.
"Ninety-five percent of this is being driven by the economy," he said. "The momentum is still there. Sooner or later we will get a U.S. OEM (original equipment manufacturer). It is just a matter of time."
Just as important, the company has edged into the corporate market, turf that Advanced Micro Devices and others have barely touched. Approximately 25 percent of the Crusoe-based notebooks went to corporations, Allen said. NEC and RLX Technologies have adopted the chip for dense "blade" servers as well. Hewlett-Packard is also tinkering with a demonstration unit from Transmeta, he added.
The interest in the company's chips comes largely from Crusoe's lower consumption of power. With wireless becoming more popular in notebooks, PC makers have to worry about conserving battery life. Embedded device manufacturers and blade server designers also need to cut down on heat, which can be reduced through energy-efficient chips.
"In our case, if lower price is all we offered, Intel would have won back every design win by charging $10 less," Allen said. "The reason almost no one has made money (competing against Intel) is that they have gone for a me-too chip."
The average selling price for Crusoe is around $100, according to most analysts. By contrast, Rise Technology, a failed Intel competitor, rarely sold its chips for above $30. AMD has tried, and failed, for years, to hit an average price of $100.
Nonetheless, Transmeta is a flea in the overall processor market. IDC estimates that the company's chips accounted for 0.3 percent of processors produced in 2000 and about 2 percent of the market for notebook processors.
"They've got a part of the market that is interesting. The question is whether they can generate the volumes that will make it all worthwhile," said Kevin Krewell, an analyst at the Microprocessor Report. "It is a slice of a slice of a slice of a pie."
The embedded market could help the company build volume, but it won't be a panacea; a large number of companies compete in the embedded sector. Other sources pointed out that companies often resort to the embedded market after being unable to sell the chips elsewhere. Embedded chips can also be quite cheap, although the volumes can be substantial.
Transmeta also will see increased competition over notebooks. Not only is Intel coming out with more low-power notebook chips on a regular basis, but Via Technologies also will more aggressively seek out major adopters of its designs. Via's chips are typically less expensive than Transmeta's.