“So where are you off to this week?” a neighbour asked. “Paris and Munich,” I replied. “Oh, very nice,” he said. Well, perhaps.
Actually, apart from the bus from the airport to the hotel, the hotel in Paris could have been a company event anywhere in the world. It was Freescale’s Paris Technical Forum, where they gather together customers and prospective customers (around 600 people in total) from across Europe and present a series of technical papers and tutorials alongside a small scale exhibition (Technology Lab) with booths from partners as well as their own product streams. It is a single vendor version of the ESC or embedded world, in essence. Since they have a bunch of senior company people present, including recently appointed CEO, Rich Beyer, the company also uses this as an opportunity to update the press and analyst community.
Apart from their “can your badge do this” competition, see www.canyourbadgedothis.com, there wasn’t really an official “Big Story,” but to a number of the journalists there was an underlying sub-text. Just to remind you, Freescale used to be Motorola Semiconductors. The separate division was first spun out into a standalone company, and, after a short period as a company whose shares were quoted on a stock exchange, it was bought by a group of investors. To pay for the shares, the investors sold a series of bonds secured on the company. These are, effectively, loans to the company which have to be repaid eventually and on which the company has to pay interest – currently around US$ 700 million a year. Using a rough figure of $100,000 for an average engineering salary in the US, that’s 7,000 engineers. OK, let us say the cost of employment is 2.5X salary — that still equates to nearly 3,000 engineers. Rich Beyer inherited this position when he was appointed, and, in conversation with journalists, he tried to make light of it. But he seems to be in a difficult position: if the company were totally privately held, then the management would answer only to the investors and could get on with the job of managing. But since Freescale’s bonds are traded, Mr Beyer has to produce quarterly earnings statements and explain significant changes in the company, just as though Freescale were public. In around five years’ time, he also has to find large amounts of cash to buy back or re-finance the bonds.
To keep the company moving, he is focusing, like so many other companies, on the businesses where he feels he can get best value, and that no longer includes chip sets for cell phones. (This from the company that in a previous incarnation was responsible for the electronics of the first practical hand-held cell phone.) While Freescale is staying in the base station business, the company is looking for a partner or purchaser for the chip set business. So if you have some spare cash – please give Mr Beyer a call.
He is also continuing Freescale’s move into creating a fab-lite strategy, running only those products that require conservative processes through the company’s manufacturing facilities and relying on foundries and membership of the IBM Common Platform for more advanced processes.
Now, since you are reading this to learn about technology and tools and stuff (at least, that’s why I think you are reading this) – why should this matter to you? Well Freescale is one of the bigger suppliers of the silicon that you use to build your systems. For example, two product areas that I discussed with Freescale people in Paris were Zigbee, or rather the whole set of wireless technologies based on IEEE 802.15.4, and a new power switching chip, effectively an intelligent and single chip relay controlling up to four sets of automobile lights. If Freescale is having to spend so much of its efforts in servicing debts, what resources will it have to invest in new products and new market areas with possible risks? The company’s policy will be to continue to concentrate on products in safer and more mature markets. This will mean that if you want to create exciting new user products for these new market areas, you will not be able to base them on Freescale chips.
NXP, previously the Philips semiconductor division, is in much the same position as Freescale, although it has actually managed to sell its wireless operations — for a reported US$1.5 billion. (NXP cost the private equity partners around US$10 billion in cash and in taking responsibility for the company’s debts.) Since the private equity investment, both Freescale and NXP have dropped out of the list of top ten chipmakers. Private equity seems to behave, at times, rather like the asset strippers of the 1970s and 1980s who bought struggling companies at a low price, sold off what they could, usually for a profit, and closed down the rest. (Not that there is any talk of closing either NXP or Freescale.)
Interestingly, as these companies are looking to focus, sell-off, and reduce costs, TI announced a new research lab, “Kilby Labs.” Jack Kilby is credited with building the first integrated circuit, just 50 years ago. He always said that the honour should be shared with Bob Noyce (although the original idea of creating an IC was first described by a Brit, Geoffrey Dummer, in 1952). The work of the Kilby Labs will be wide ranging, but it will all be IC centred. By one of those coincidences that editors like, the same week that Kilby Labs was created, it was rumoured that Bell Labs, now owned by Lucent-Alcatel, will be “pulling out of basic science, material physics, and semiconductor research and will instead be focusing on more immediately marketable areas.” That’s the Bell Labs that gave us the transistor, Shannon’s information theory, Unix and C, and much more.
But back to last week. Munich, the day after Paris, was different. PR agency, Publitek, had rented space in a resort hotel by a beautiful lake in a defiantly picturesque village in the mountains to the south of Munich. Four companies showed off what they would be launching at electronika, the huge German trade show in mid-November. Much of what was discussed is under embargo (i.e. I can’t write about it) until the show opens, but look out for interesting announcements from ceramic specialist, Murata; USB company, Future Technology Devices; mixed signal company, Silicon Laboratories, and analog specialist, Intersil. (If you are visiting elektronika, Intersil is promising Analoger Beer and T-shirts on their booth, while Murata will be showing robot demonstrators.) The pre-electronika briefing was attended by journalists from all over Europe, and a lot of gossip about the electronics industry changed hands in the social sessions, much of it sadly un-reportable, but some of the gossip may guide what I write in these columns over the next few months.
The week before last was the UK’s annual Embedded Systems Show (ESS). Normally this is another source for product announcements and industry news, but this year, news was a bit light. (There were some sessions where the result of company briefings will see the light of your screen later this year.) While the show was taking place, the economic tsunami was continuing to build, but the exhibiting companies were not seeing serious indicators of recession among their customers. A project has slipped a few weeks here, a person leaving is not being replaced there, but so far no reported project cancellations or major lay-offs such as we saw in 2001/2.
What was also clear at ESS is that the software development tool suppliers are still having an up-hill struggle to persuade medium and small embedded development teams to adopt the tools that will give them the ability to build better products in shorter time scales. But I will be returning to this theme again later in the year.