
EDN Europe's Editor Graham Prophet posts a selection of comments and insights prompted by the many items of industry news and rumour that cross the editorial desk or are gathered on his frequent travels to interviews, press conferences and events around Europe - and further afield - and somehow never find their way to the
magazine or the web site, recovering some of the information otherwise lost in the noise level...
Friday, June 05, 2009
More Intel
I wrote recently of the mixed experience I have had over the years of being on the receiving end of Intel’s marketing and promotion output – especially as that relates to Intel’s involvement in the embedded market space. It’s been….. “variable” is a good word; there have been times when, as an outside observer, you’d never guess that Intel sold into the embedded space at all: and there have been intermissions when (like now) the company has suddenly rediscovered the embedded space, and noticed all over again that it’s a major player in it.
Now, it has dipped into its savings-and-loan thrift account and bought Wind River for over $800million. What’s the point of that deal? If you wanted to coin a single phrase, then “virtualisation in the embedded space” might not be too far off the mark.
Back up a little. I just spend a couple of days at a technology review meeting, and 25th birthday party, of the research organisation IMEC (more of that elsewhere). In the course of which there was, naturally, much discussion of (the terms already have cliché status) “More than Moore” – that is, keeping the scaling that underpins Moore’s Law going – and “More than Moore” – the added aspects beyond process scaling that will keep the semiconductor and related businesses vibrant in the coming years. One of the presentations was by Intel’s senior vice president Bill Holt.
Holt re-emphasised some of the fundamentals of the industry, including the fact that it is the constantly falling cost-per-transistor that functions as a primary driver. Far from asking if Intel can see an end to semiconductor process scaling, Holt clarified the fact that Intel can’t afford for the pace of that scaling to slow. So Intel is going to continue building products with more and more transistors.
We’ve already seen that even Intel is finding it tough to squeeze more transistors into a single processor; so, they stack up the processors and we get multicore. Over time, multicore products become the standard offering; Intel, alert to the ubiquitous-computing future places renewed emphasis on its embedded-systems markets; buying an OS-and-infrastructure product line, and one that knows a bit about handling multicore in the embedded space, looks pretty logical. In fact, part of Holt’s presentation talked of the “need to improve synchronisation between process and design” – tightening the links right along the product development chain. When he said that, of course, he probably knew he was on the point of closing the deal to buy an OS company – his audience didn’t.
As I said here a few days ago about the TI/Luminary deal, hard times are a good opportunity to acquire stuff if you’ve got the money (note that the WR takeover is for cash, not paper). When the going gets tough, the tough – or, at least, the well-heeled – go shopping. This one won’t be the last big takeover, by any means.
Post a comment
Note: fields with an asterisk(*) are required information.
All submissions are subject to review before they are posted live.