It looks like Déjà vu all over again – Yogi Berra

OK. Every meeting I go to I’m asked my opinion regarding Intel’s acquisition of Wind River – so here goes. To be honest, people richer than me are behind and part of this acquisition, so who am I to think that I know better? To be sure, some of the people richer than I am have screwed up on a far grander scale than I have, so why not offer my perspective?

My first response to the question is a flashback to the “Shortest Book” jokes that I used to revel in decades ago. “Ethics for Enron Executives” would certainly be among the “shortest book titles”. My contribution to this arcane joke file was “Famous Jewish Weightlifters” a title that my rabbi didn’t understand.

Recently I thought of adding “Successful Intel Acquisitions and Spin-offs” to the list. Those of you old enough to remember Dialogic and Ziatech (as well as a number of software-based hardware analysis and virtual development platform spin-offs) will get my point.

So what’s in it for Intel and why did Wind River bail out so easily? Let’s look at the facts:

• Intel doesn’t need the additional $300 million on their balance sheet
• Wind River has been losing market share for some time now and their “subscription model” (which they have abandoned) cost them dearly in the mil/aero marketplace and allowed Green Hills and LynuxWorks to significantly underbid them
• Rob Davidson, Wind River’s VP of A&D, has publicly stated (at an Intel sponsored event no less) “let me make it clear that Wind River is a Freescale company” since 85% of Wind River’s military sales are Freescale-based
• Intel, notwithstanding their successes, is facing considerable competition from ARM and FPGA vendors – particularly Xilinx
• The Mil/Aero market and the primes that spend the most on embedded technologies and sub-systems are decidedly moving to a “software centric” purchasing model

Has either of them thought through their respective channel strategies? Remember ages ago when Wind River bought out their arch rival ISI and acquired pSOS? The bloodletting that followed was inevitable given the incompatibility of their respective sales forces. A funny thing has since happened to Wind River, who appeared at that time to monopolize the embedded OS marketplace. Has Intel anything to lose by acquiring Wind River – will they incrementally increase their market share and influence within the embedded world.

Twenty years ago Intel and Microsoft ruled the embedded world (they were referred to as WinTel). Today, the parameters of the marketplace have changed significantly and here we see a hardware company attempting to insure their dominance by acquiring a high-profile software company. Had they merely wanted to increase their product offerings Intel could have bought LynuxWorks at a fraction of the cost of the Wind River acquisition.

Truth be told – both Intel and Wind River are missing the emergence of an embedded market that will soon overshadow the rest of the marketplace. It involves several components – none of which are within the new Intel’s capabilities.

• Systems development – with C and C++ developers working simultaneously on the same development
• Unique problems associated with systems-within-systems developments and long term support of such
• The new era of “Smart Devices” – the merging of IT database capabilities securely communicating with intelligent embedded devices

While Wind River, Green Hills and LynuxWorks are fighting for the limited and misunderstood mil/aero markets that demand DO-178B and MILS level security, operating systems such as ThreadX, Nucleus and Micrium along with MontaVista Linux are providing significant competition – particularly for consumer electronic and medical markets in which the selection of OS and processor to best match the product requirement is displacing the FUD surrounding the “suggested” need for excessively stringent security. ThreadX, Nucleus and MontaVista have hundreds of millions of applications running world wide and Micrium is already DO-178B certified.

EMF ROI analysis shows how these lesser known OSes are outperforming their larger and better known competitors. We don’t see how this helps Intel in the long run.

For the past 5 years, Wind River didn’t consider itself an embedded company but rather an enterprise company. This was a silly way to position themselves and they have paid the price for such folly. Embedded requirements far exceed the 5-nines requirements of enterprise developments.

Finally, over the ensuing three years, the DoD will reduce discretionary funding by between 18% and 40% depending on which reliable source one is using. This will create a huge hole in mil/aero OS markets and will seriously impact Wind River.  Maybe they went for the acquisition to come under Intel’s huge financial umbrella.