Reproduced by kind permission, Peter White, Rethink Technology Research Ltd.

There is something scary about Simon Atkinson at Mirics Semiconductor in the UK, you can tell. I think it’s because he can see the future. He came out of the RF side of the ancient military communications company Plessey, worked at LSI Logic for a while, then started his Mosaic Microsystems business (no not the browser company, a chip design house), which when sold to Analog Devices eventually formed a huge part of cellular RF business, which Atkinson went on to run.

That business, at the time was supplying cellular chips to Siemens handsets, later acquired (and then closed) by BenQ. Eventually the Analog Devices unit was sold in 2007, mostly for its head start in TS-SCDMA chips for the Chinese market, with Taiwan chip designer MediaTek paying $350 million for it.

Someone at Analog Devices would have sighed a sigh of relief, glad to be shot of Atkinson. This is because he’s one of those people who can see exactly what needs to be done next in any given company, and embarks upon it whether or not the his boss agrees, simply because it’s the best thing for the company.

Which is how come the ADI unit had developed a software defined radio as early as 2003, which would eventually be able to handle both WCDMA and TD-SCDMA in the same radio – something that has only recently been made commercially viable at mobile chip leader Qualcomm, which now has a chip that can work with both W-CDMA and CDMA2000, as well as GPS.

The scary thing about Atkinson is that he is fascinated by all the “big” problems in the world of radio, and for some considerable time he has seen the issue of a software defined radio as his personal bête noire.

Inside ADI, it took the form of powering its Othello line of cellular chips, and offering a bridge between two different cellular architectures in different areas of the spectrum, while being dynamically re-configurable between them. Another chip he had back at ADI offered the same kind of bridge between 2G and 3G. So why didn’t it catch on?

“People didn’t even realize that 3G was even going to catch on back then, they were worried about the day to day problems of cellular. The world of cellular transceivers was a brutal world and it was all about security of supply lines, not innovation,” Atkinson tells us.

A purely corporate manager, in that kind of environment, would have had his hands full with Atkinson, who only wanted to look five years out, calmly see what to him was the obvious future, and then equally calmly go about the business of intercepting it.

It was then that he realized that his passion for software defined radios needed to be taken into an environment where he was solely in charge – and raised the funding for Mirics. We can see the ADI point, even if Atkinson can’t. Software defined radios had cost companies such as Intel $billions in R&D, without any successful products. There was a pointed comment when the unit was sold to MediaTek, which said that now it was in a home which could afford the R&D needed to make the technology behind these products work. We
figure it was aimed personally at Atkinson, but he doesn’t care.

“Mirics is about taking on the same problem but in the much simpler broadcast environment. We are not a company set up to produce just one product, but we shall learn more and more about wireless convergence with each product that we come up with, and our long term aims are about converging different content delivery types onto a single chip platform,” explains Atkinson.

Mirics’ first product was simply a radio tuner which worked up and down the spectrum and was sold to other companies to sit alongside their demodulator products.

The MSI001 chip was capable of receiving all broadcast and could work with DVB-H, T-DMB, ISDB-T, DAB-IP, MediaFlo, DAB, DRM and even AM and FM demodulators, as well as unknown future standards, because it was designed to be a co-processor for the Standard’s demodulator chip. The device provided coverage from 100KHz to 1.9GHz and mostly it was used for DAB, FM and AM.

But then, in 2007, Atkinson saw the future once again, unfortunately it was just before his company had run out of money. The TV market was fragmented and likely to become more and more fragmented, and TV was going to move more and more onto laptops, PCs and mobile phones. He then convinced his team, in just two months, to create software based demodulators to run on an Intel Pentium and grabbed another $10 million from Intel Capital, Pond Ventures and Acacia Group.

His problem, he had realized, was that by supplying a component to other companies which controlled the route to market, he was at the wrong end of the food chain. All they did, he reasoned, was to build a demodulator, so why couldn’t he build one, and then move up the food chain.

If you ever meet him over a beer, let him tell you how him and his team managed to get a working prototype out of the door before the investors came around, but for now, two years later Atkinson and Mirics is on the verge of turning that work and that decision into a commercial success.

The outcome is FlexiTV and by putting the demodulation, that part which defines the exact radio format, into Pentium driven software, it has opened up a whole raft of applications for Mirics. We first came across the company when it had just managed this in September 2008, and since then it has raised yet more money, put DVB-T onto PCs and Notebooks, built a version of its software to run on the underpowered Intel Atom which lives in most Netbooks; partnered with Chinese chip designer Spreadtrum to make a CMMB demodulator on PCs, and developed a version of FlexiTV that will run on the ARM Cortex 8 processor which drives devices like the forthcoming Nokia N900 and the Sony Ericsson Aino and Kurara.

It has also partnered High Definition Digital Technology Industrial Company (HDIC) in Shanghai to come up with what is only the third reference platform for the nascent digital terrestrial standard in China, the one which was once called DMB-T/H and is now referred to as Digital Television Terrestrial Broadcast (DTTB).

Only Legend, the Sino-US company that designed the TDS technology behind the service, and Microtune who acquired Auvitek in Shanghai, also have products which serve this market.

“DTTB is a bit like ATSC in the US, and it’s now deployed in 76 cities in China, so don’t think digital terrestrial in China has gone to sleep either. Unlike the mobile TV service CMMB, content is offered free to air and there are 5 HDTV channels already live.”

So when he says, “I can’t talk about who we have lined up a deal with, mostly because it is not completely closed and I’d feel a fool if it all fell through, but we are on the verge of shipping an awful lot of our FlexiTV product,” we are left guessing whether he is talking about one of the major PC manufacturers with a global footprint, one of the leading handset vendors or a CE manufacturer that wants to target Chinese TV set tops. So that’s Dell or HP,
Samsung or Sony potentially buying into this concept.

When Atkinson pulls out the USB version of his FlexiTV tuner it looks tiny, smaller than a HSDPA dongle, and is has a Bill of Materials less than $5, and in large volumes that can clearly go down considerably.

“Just look at the way Freeview and DVB-T has worked in the UK and across all of Europe. Everyone has bought HDTV ready for High Definition and a new multiplex using DVB-T2 has come along in the UK, and it will probably go the same way in the rest of Europe. All those people who have bought brand new digital TVs will not be able to view the DVB-T2 channels, and will be on the look out for a new converter box.”

“Wouldn’t it be better to have high quality dumb screens that work with any input and a single device which can tune a DVB-S or DVB-S2 satellite signal, or a DVB-C cable signal or a DVB-T or DVB-T2 terrestrial signal and anything else you like. That’s where we are going, making a chip that can do all of that on the same piece of silicon, software configurable. I’m not saying we are there now, but it’s where we are headed.” Told you he was scary.

Faultline is published by Rethink Technology Research Ltd, 1 Wide Lane Close Brockenhurst, SO42 7TU
Tel: +44 (0)1590 624530 Fax: +44 (0)207 900 2225

Faultline Principal Analyst: Peter White peter@rethinkresearch.biz

About Rethink
Rethink is a thought leader in quadruple play and emerging wireless technologies. It offers consulting, advisory services, research papers, plus two weekly research services; Wireless Watch which has become a major influence among leading wireless operators and equipment makers, and which has pioneered research coverage of WiMAX; and Faultline, the Journal of Quadruple play Economics, which has become required reading for anyone
operating in and around quad and triple play services and digital media.