Blog Post

Why IBM, Chipmakers Are Backing an Open-source Linux Company. Hint: Intel

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

ARM (s ARM), IBM (s IBM), ST Ericsson, Samsung, Freescale and Texas Instruments (S TXN) are banding together to back a not-for-profit open-source company, Linaro, which will develop tools, packages, processes and a Linux kernel (aka a core component of the OS) that will run on any system-on-a-chip currently used by non PC-devices. They will fund the UK-based company to the tune of tens of millions of dollars and will liberally spend money on hiring crack open-source engineers — nearly a 100 of them. And their target is none other than world’s largest chipmaker: Intel Corp. (s INTC)

The idea behind Linaro is to develop a Linux-based set of tools and software that can be used by anyone, from folks who make operating systems for smartphones and tablets to set-top boxes. It’s amazing when you start to think about how many of our consumer electronics devices and newer forms of computers are running on a variant of Linux.

What that means is that there’s an increasing diversity of operating systems and chips they run on, and this complexity is only going to increase as more always-on, connected devices come to market. At present, most of the chip vendors (and hardware makers) do a lot of the same work internally, costing them a lot of money, argued Tom Lantzsch, executive vice president at ARM.

“We are at a critical inflection point and it is becoming clear that most of the large Linux-based efforts want to scale to multiple form factors,” said Lantzsch in an interview earlier today. The company will release software on a six-month basis, which would in turn allow companies to focus all their energies on developing user interfaces, applications and other enhancements to their operating systems. And because of the open source nature of the effort, theoretically speaking, everyone from Android to MeeGo can use Linaro’s software and enhance it.

Click on the image for a larger view

In reality, however, I don’t see that happening, given the tight integration Google has done on the Android OS. Still, there is a merit to the idea -– with an increasing number of consumer devices becoming digitized and connected, there needs to be a semblance of order at the very basic software level. When I asked Lantzsch who would be impacted by this new effort, he gave me a politically correct answer. He dismissed the idea that this new effort would impact MeeGo, Ubuntu, Android or anyone else.

I suspect the real target here is Intel. If you guys remember, last year Intel bought Wind River Systems, a company that had a virtual lock on the embedded systems and related software, for $884 million. With this new effort, these big companies are trying to do an end run around Intel and its Wind River division. This business gives Intel an advantage when trying to push its low-power Atom processors, which coincidentally compete with ARM-based processors.

In her analysis of the deal, Stacey wrote:

All sorts of electronics (yes, even your toaster) are growing more complicated, and more of them are also are getting connected to the web (not yet on the toaster), which means they are able to send and receive data. Once that happens, they need bigger brains to manage and react to that flow of information. Hence the need for smarter embedded chips, and Intel’s efforts to take its brand of x86 general-purpose computing to all devices.

The purchase of Wind River also gives Intel a customer base that includes the automotive, consumer electronics and industrial companies to which it wants to sell its embedded chips, giving it access to the customer base that buys embedded chips. Intel’s move into the market will be highly contested by the likes of Texas Instruments, Freescale and ARM, which are already selling into the embedded and wireless spaces.

It has taken Intel’s rivals a year to respond -– and the response is called Linaro. Let the battle begin.

8 Responses to “Why IBM, Chipmakers Are Backing an Open-source Linux Company. Hint: Intel”

  1. I am impressed with your analysis that links between Linux and Intel, we know that the intel processor is the king of the world, if it could join Which linux is open source I thought it would be a big hit in device development.
    IBM has a long experience in the field will certainly take advantage of Linux servers and Intel to make a super storage and servers that have a high sale value

  2. lc386

    We should ask one think before jump to the conclusion today…lots use Linux on x86 in the past but ARM partners shipped 4B chips (1.4B phone) annually and Intel/AMD x86 has around 300M (PC/laptop/server) today. How many Linux OS running on ARM and how many on x86 for production today? Don’t underestimate the embedded world.

  3. Om, frankly this is no more than yet another flash in a pan. A random mosaic of competitors and de-riskers (hint: oodles of unclear strategy + poor commitment at display) trying yet again to de-fragment Linux with mobile phones in mind. Remember CELF (CE Linux Forum)? Look at embedded alley (EA), monta vista at a plethora of linux merchants. Look also at MeeGo and moblin and several other “backyard” linuxes.

    Making a CE product entails a huge amount of technical complexity. And, most often, the OS is not the biggest piece of the puzzle. Face it – an OS is commodity. To be able to sell your product, you need differentiating features. And that calls for differentiating investments. Not a sub-$100 million investment in yet-another-commodity Linux packager.

    Finally, why do you point at Intel? OS initiatives and vendors come and go as surely as seasons. ARM based SoCs rule the phone world. Intel is not a player a phone market. And there are hardly any non-wintel player left in the netbook market. I guess this peace will last a while – Intel will focus on netbooks apart from their near monopoly in laptops and PCs. ARM and MIPS based SoCs will rule the mobile phone and CE worlds (TV, STB, BluRay,MP3, ..). Since Android doesn’t yet make sense on non-mobile phone CE devices, naked Linux distros will thrive there. Linaro and its ilk could sense an opportunity there, but it’s a far far stretch of imagination to ink dots all the way to intel.