•  
• EDN Europe MAGAZINE
•  
• Free SUBSCRIPTION
• Subscription Renewal
• View Media Kit

• EDN
• EDN Australia
• EDN Asia
• EDN China
• EDN Japan

EDN.COMMENT: TI/Luminary, Intel/Wind River— it’s not over yet

Yet more M&A (merger-and-acquisition) activity in our industry, although in these latest examples—unlike the forthcoming marriage of Renesas and NEC (Reference 1)—the “M” part of the acronym is not very relevant, these being straightforward takeovers. Texas Instruments has snapped up Luminary Micro, adding the Stellaris (ARM-Cortex-based) product line to its 16-bit MSP430 and high-end C2000 ranges. No surprise to see a venture such as Luminary being bought; something of a surprise, perhaps, to see it being bought at a relatively early stage in its development. When you buy a company that has made a point of saying, “Don’t mess with 8- and 16-bit chips any longer, our 32-bit parts will do all that!” (my summary, not their words), how do you position the future of your own 16-bit product line?

In answer to just that question, a TI spokesman told me that the MSP 430 product line would continue to develop, especially in the direction of everlower- power offerings. Emerging market sectors such as smart metering, he said, will require significant processing power while demanding very low current drain. In the Cortex MCU sector, this takeover adds further weight to the array of microcontroller companies offering M3-based products and, if ARM’s CMSIS interface-standard initiative delivers on its promises, increases the prospect of a very hotly contested market space.

More recently still, Intel has dipped into its savings-and-loan thrift account and bought Wind River for over $800 million. What’s the point of that deal?

One thing we can confidently expect of Intel is that over time, it will continue to build more and more transistors. In a recent presentation, the company’s senior vice president Bill Holt revisited some of the fundamentals of the IC industry, including the fact that a constantly falling cost per transistor functions as a primary driving force.

We’ve already seen that even Intel is finding it tough to squeeze more transistors into a single processor; consequently, it multiplies the processors and we get multicore products. Over time, multicore devices will become the standard offering, and designers in the embedded space will have to become adept at handling them. Intel, alert to the ubiquitous- computing future, is currently placing renewed emphasis on its embedded- systems markets: buying an OSand- infrastructure product line, from an enterprise that knows more than a little about handling multicore in the embedded space, looks pretty logical.

When the going gets tough, the tough—or, at least, those with cash available—go shopping. It doesn’t seem likely that these deals will be the last takeovers we see in the present round.