But that situation will begin to change this year, as cheaper and better-designed Bluetooth components begin to appear in volume, according to a British company at the forefront of Bluetooth development.
Cambridge Silicon Radio (CSR) has been attracting attention for its single-chip Bluetooth implementation, which uses standard silicon and could allow Bluetooth to be added to devices such as mobile phones and personal digital assistants (PDAs) for just $5. Most recently Sony showed its enthusiasm for CSR's approach by investing in the company, to complete a $22 million round of funding.
The company, which has been showing off products this week at the Consumer Electronics Show (CES), believes Bluetooth is just about on track, despite the sense of disappointment. "With Bluetooth, the time has been maybe only about six months slower than anybody predicted one or two years ago," says Phil O'Donovan, CSR's managing director and cofounder. "It was the same with GSM. People kept saying it's about to take off, it's about to happen. It finally took about three years."
Industry observers agree expectations have been distorted by the early hype, which began at the end of 1999 when Bluetooth first caught the public imagination. "It is really going to happen," says Mat Hanrahan, analyst with Bloor Research. "It only seems slow because of the hype. You don't have to know much about technology to talk about [Bluetooth] or envision a Bluetooth-enabled world. It's been let out of the bag far too quickly."
In conception, Bluetooth is simple, which may be one reason why the process of bringing it to products seems so slow. Radio chips are embedded into products ranging from refrigerators to laptop computers to mobile phones, allowing them to connect to one another when in the area and exchange data. The killer applications are expected to be, for example, the ability to connect your laptop to the Internet via your mobile phone, without having to line up the infrared ports or even take the phone out of your briefcase.
But making this all work seamlessly is another matter, with the two immediate hurdles being cost and interoperability. You can buy a Bluetooth card for your laptop today from Toshiba, and get a Bluetooth headset for your Ericsson mobile phone, but the devices will cost hundreds of pounds and the gadgets may not all recognise one another.
New chips such as CSR's BlueCore01 are aiming to solve the pricing problem. Experts and industry players agree Bluetooth chips must reach the $5 range to begin to be economical, but most today cost at least $15.
CSR claims to be the first to have integrated all the Bluetooth components onto a single chip of industry-standard silicon, greatly reducing manufacturing costs. The company has sold more than 100,000 of the chips for $8, and says it will crack the $5 barrier late this year.
A San Diego, California company called Silicon Wave also claims to have a cheap chip design, and chip companies such as Conexant Systems, Broadcom (through its acquisition of Innovent), Atmel and National Semiconductor are also competing for the Bluetooth prize.
There is a lot to be gained: Merrill Lynch predicts Bluetooth sales will reach $4.3 billion by 2005, and only the elite will be able to carve up that market. "Today there are 25 to 30 providers, and two years from now there will be a lot less," says Ron Dennis, Merrill Lynch semiconductor analyst. "The first to market clearly has an advantage here."
The other important factor, interoperability, got a boost late last year with the formulation of version 1.1 of the Bluetooth specification, according to Dennis. The spec is due for final approval early this year. "Interoperability is key," Dennis says. "In order for it to work and work right, the spec has to be standardised. Every product has to work with every other product."
Dennis expects more Bluetooth chip manufacturers to begin volume production in the second or third quarters. But even after the silicon is out on the market, don't expect the world to be Bluetooth-ised right away. For one thing, there's still the software to be ironed out. "There's not a lot of software around, and there needs to be a lot more," admits CSR's O'Donovan.