File Transfer

Meet the man who named (and helped make) Bluetooth

If you’re reading this, you’ve probably used Bluetooth, the short-range radio connection that acts in place of a wire to connect phones, headsets, laptops and other devices. But the story of how it was created is one of an all too rare victory for consensus in a technology industry where politics and stubbornness can drain the life out of even the smartest ideas.

If marketing had had its way Bluetooth would have been called PAN, short for personal area network. That wouldn’t have been a bad name: the prefix ‘pan’ has associations with universal connections and in network terms a personal area network is exactly what Bluetooth creates. But the name was already trademarked so an Intel engineer and amateur history buff called Jim Kardach who had been a leading figure on the project got his pick. (Bluetooth was a King of Denmark and Kardach had come upon the name when reading a book about the Vikings.) Recently I spoke to Kardach for an hour by phone and he told me the story of how Bluetooth started to become the pervasive presence it is today.

“I was in the mobile microprocessor group at Intel,” he recalls over the transatlantic phone line. “I’m a low-power guy and back then I had maybe 50 or 60 patents on things that end up in your notebooks.”

In 1997, Kardach had just finished his work on ACPI, a specification that lets the operating system control power management hardware in computers, and was asked to figure out how notebooks could get onto the internet wirelessly. 

WiFi, the commercial implementation of the 802.11 spec, was not the staple of wireless infrastructure it is today. Nor was there much in the way of universal agreement when it came to cellular network links.


Broken connections

“802.11 wasn’t a success at that time: it cost $2,000 to implement and it barely worked,” Kardach says. “If you had cellular, there were no standards: America had three different standards, Japan had its own thing, Europe was using GSM… so we were looking at a way people could get onto the internet with their notebooks. Everything was too expensive. Back then you’d connect your notebook to your phone with a wire and we thought maybe we could make it cheaper or create ‘a wireless wire’.

“One of the problems with notebooks getting on the internet were all these physical cables. Back then you would have a kit for all the different plugs, a phone plug for each country, to allow you to connect the phone line to your modem. Us US guys could burn out our modems if we accidentally plugged into a European ISDN socket. We were looking at all kinds of things. I was looking at things that would work everywhere including satellites because Iridium was going to do that.”

It was against this backdrop that Kardach was put forward by Intel to get involved in what became the Bluetooth project.

Ericsson’s MC Link technology, designed to link to accessories, came to form the basis of the Bluetooth Special Interest group (SIG) and Ericsson, Intel, Toshiba, IBM, Motorola and Nokia assembled to form a working group.

“There was a MoU [Memorandum of Understanding] in 1997 and we all met up in Lund, Sweden and agreed on this,” Kardach recalls.

However, a notable no-show was Microsoft which made what was to be the common mistake of thinking that Bluetooth could derail what was to become WiFi - more of which later.

Go faster

Kardach says that the remarkable swiftness with which Bluetooth developed could be put down largely to the group having a clear set of goals – the approach was lined up early on. Challengingly, the group wanted a product built on a standard bulk CMOS process that would scale along the lines of an Intel or other popular microprocessor. Also the package had to cost OEMs no more than $5, be secure and be built on the 2.4GHz ISM band so that it could work globally.

“We had made a lot of decisions upfront and the reasons I think it went smoothly had a lot to do with the SIG’s IP policy which meant that members of the SIG could use the necessary IP of other members freely. This made a very large difference in how engineers interacted with each other; instead of pushing technologies because a member’s company might get potential revenue when the technology took off, engineers would pick technology or directions based on merits in meeting our goals. It may sound silly but we picked our goals very early. It had to work around the world and it had to be secure so when we were arguing it was about how to meet those goals. If you’re arguing about the goals themselves you’ve got problems and you’ve lost your direction.”

The spec moved on apace.

“Some of that goes back to Intel being an execution machine,” Kardach believes. “It was driven hard like any [Intel] programme and we started with good technology like Ericsson’s MC Link and elements from the Motorola Piano service discovery protocol. The difficult thing was we had to put it together into an end-to-end system. But we had extremely brilliant people from all over the world helping us to define these radios. I remember the names of Jaap Haartsen (father of the baseband), Sven Mattisson (father of the CMOS radio), but there were many others from Nokia, CSR and others.”

There were still challenges of course, notably the lack of true harmonisation on the 2.4GHz spectrum that had been adopted which meant that work had to be done on a country-by-country basis. Still, in the first half of 1998 a media launch took place in London and Bluetooth became a word on the lips of techies the world over.

With the benefit of hindsight, Kardach says that the 1.0 spec that emerged in June 1998 was “probably a bit premature and had early interoperability problems, resulting in new ‘dot’ versions quickly following”.

The early radios were more expensive, touching perhaps $25 but that bulk CMOS commitment meant that Moore’s Law kicked in and process improvements led to lower costs. Today, adding Bluetooth to equipment costs under one US dollar.

‘Something for everyone to hate’

At the time there were still disputes about Bluetooth versus 802.11, despite Bluetooth being always intended for dense, short-range connections. In fact, Kardach says, the Bluetooth camp spent a lot of time attempting to ensure that Bluetooth and 802.11 would work together as good citizens via frequency hopping to avoid clashes, for example. Adding to the misunderstandings that seem to go hand in hand with the introduction of any new and disruptive technology there was also confusion over another networking effort for domestic use called HomeRF.

“A lot of people couldn’t understand the difference between a PAN and a LAN but to us it was very different,” Kardach says. “Even within Intel we were having huge fights with the WiFi guys. Ultimately you’re all passing data over radio waves but the details were more technical and the people who were technical got it. But yes, for a while there was something for everyone to hate.”

Bluetooth’s success also came quickly, even if it was at first disconcerting to see users appearing to be talking to themselves when using Bluetooth earpieces or hands-free kits. But Kardach still laments that some developers applied Bluetooth to tasks for which it wasn’t suited – including, yes, LANs.

“A lot of people wanted to use it for a lot of applications that much later Bluetooth LE [low energy] was really targeted for,” Kardach says today.

Fame of a kind

Bluetooth wasn’t perfect and there were early issues with crashing equipment but such was its success that the engineers associated with Bluetooth became celebrities of a kind.

“Everyone working on Bluetooth became famous and it was bit uncomfortable sometimes,” Kardach says, adding that at a Middle East university he was “mobbed like a rock star by 300 people asking for my autograph”.

The recognition must have been nice after all the hard work, surely?

“I’m a bit shyer than most people… I didn’t dislike it. Bluetooth probably went beyond even what I thought - and I thought it was going to be the biggest thing since sliced bread. And in many ways it was.”

Kardach is clearly proud of his work on Bluetooth and on projects like ACPI, SMI, the Intel 386SL processor and the same company’s Haswell microarchitecture.

Today he’s no longer at Intel, which he retired from after a 26-year run, but he’s still working on saving electricity and the never-ending pursuit of efficiency at FINsix, a startup that develops high-frequency power conversion for very small computer power adaptors.

“I’m a low-power guy,” he repeats, modestly.


« Crowdsourcing Innovation: Aidan Chopra, Bitsbox


Ken Banks interview: Technology for social good »
Martin Veitch

Martin Veitch is Contributing Editor for IDG Connect

  • twt
  • twt
  • Mail


Do you think your smartphone is making you a workaholic?