On Tuesday, Semiconductor firm Cavium Networks announced that it had signed a definitive agreement to acquire embedded Linux pioneer MontaVista Software for $50 million. After the acquisition wraps up in December, the privately held MontaVista will run as a separate operating unit, retain its own brand name, and support multiple architectures, said Cavium. This message of independence was reinforced by MontaVista execs Jim Ready and Dan Cauchy in an interview with LinuxDevices.

Cavium, a U.S.-based semiconductor firm that focuses primarily on MIPS-based carrier-grade networking processors, is following in the lead of fellow semiconductor firm Intel, which acquired Wind River this summer for $884 million. For years, the publicly held Wind River has been MontaVista's chief rival in commercial embedded Linux operating systems (OSes) and tools, and was recently projected by VDC Research to have moved past market leader MontaVista in revenues.

The consolidation trend was reinforced this summer when RTOS firm Mentor Graphics acquired Embedded Alley, a much smaller, but up-and-coming player in embedded Linux and Android development.

Lyman: Linux consolidation now fully underway

According to a blog by Jay Lyman on 451 CAOS Theory, a blog site hosted by IT analyst agency The 451 Group, the MontaVista acquisition is a sign that "larger hardware players are envisioning a future full of embedded Linux." Lyman continues, "The best way for them to get there, it seems, is to partner with and acquire smaller embedded Linux software, services and specialty shops."

One of the key strengths of Linux, at least from the point of view of a chip company, is breadth, says Lyman. Linux is said to be strong in telecommunications, military and aerospace, industrial control, and consumer devices, among other areas. In the consumer devices, it has grown to be a major contender in smartphones to automotive infotainment systems, he adds.

The recent trend of investments by Intel, Mentor Graphics, and Cavium "is further validation that Linux, in its various forms, is the right fit for most if not all of these key embedded software markets," writes Lyman, adding that more acquisitions of embedded Linux player can be expected.

Weinberg: Why MontaVista lost ground

Bill Weinberg (pictured at right), writing on the LinuxPundit blog of his similarly named tech consulting agency, agrees that more consolidation is coming. Yet, he notes that the main targets have now been spoken for. An embedded Linux veteran who recently accepted a position heading up a new "Mobile Open Source Practice" launched by the Olliance Group, Weinberg once worked at MontaVista, and more recently consulted with Embedded Alley before it was picked off by Mentor Graphics.

According to Weinberg, the remaining potential acquisition targets in embedded Linux are much smaller than MontaVista. These include the pseudo-DIY Linux development service Timesys, the industrial- and real-time Linux focused German services provider Denx, and Linux tools supplier Viosoft, says Weinberg.

According to Weinberg, MontaVista is a good deal for Cavium for many of the reasons stated by Lyman, writing that "they get revenue growth, enabling technology, expertise and new ecosystem reach." Yet, he also suggests that Cavium may have been concerned they might not receive top-of-the-line service from an Intel-owned Wind River, the company's primary embedded Linux partner in recent years.

Weinberg notes that MontaVista has been rumored to have been courting suitors for three years now, but suggests that recent missteps have led to the company receiving a lower price than had once been anticipated. After lauding the company for its incredible rise to dominance, and for bringing embedded Linux to the forefront, a history that was detailed in our tenth-anniversary coverage this summer, Weinberg details several MontaVista strategies and decisions that have seen it lose market share to Wind River and others.

First, as an integrator of open source software (OSS), MontaVista was "often swimming below the ever-rising open source value line," writes Weinberg. "True innovation emerged from the company, but always so low in the stack (mostly in the kernel) that they were unsuccessful in commanding a premium for it."

Furthermore, the company's revenue growth was "always limited by their ability to capture development teams as customers," Weinberg continues. "They resisted both developing deployment IP or reselling run-time technology from 3rd parties, limiting their opportunity to benefit from successful high-volume OEM customer products."

Weinberg also knocks the company for a strategy that resulted in "a product line bloated with board support packages that never saw the light of day in shipping OEM products but added substantial time and costs to new releases and sustaining engineering."

While noting that some have suggested that MontaVista's business model of integrating OSS software, as opposed to complementing the platform with proprietary IP, was "essentially flawed," he argues that the real problem was execution.

Weinberg claims that MontaVista Linux releases were increasingly infrequent and "did not closely track ongoing Linux kernel and other OSS project evolution." In addition, the company lost most of its "hallmark on-staff project maintainers," and "never made sorely needed investments in truly original differentiating technologies and products."

(One might argue that latter point, especially with recent products like the Montebello fast-boot technology underlying Dell's Latitude ON, and the recent, radically reframed MontaVista Linux 6, but prior to these products, there was indeed a fairly long gap of business as usual.)

Neither Lyman or Weinberg speculate much on what the recent acquisitions will mean for the embedded Linux industry. Yet, with embedded OSS development focus increasingly moving to the higher layers of the stack, especially in the mobile device world with open source projects like Android and Moblin, perhaps it will not make that much impact at all. While Cavium's MontaVista, Intel's Wind River, and Mentor's Embedded Alley are all major players in these upper layer platforms, none of them call the shots in this more heterogeneous world.

Availability

The 451 CAOS Theory blog by Dan Lyman may be found here, and the LinuxPundit blog by Bill Weinberg should be here.

• Semiconductor vendor to acquire MontaVista
• Linux distro boasts CGL 4.0 compliance for MIPS
• MontaVista's market-specific distros support Android, Atom
• Linux-ready MIPS64 SoCs jump to 32 cores
• ARM11 SoC boasts under one-Watt efficiency
• MontaVista ships MVL6 on five platforms
• MontaVista turns ten
• Wind River's Linux chief: I don't wear an Intel badge.
• Intel's Wind River subsidiary claims embedded Linux dominance
• MontaVista touts Android services
• MontaVista rev's carrier grade distro for 4G
• Network processors double performance, run Linux
• Embedded Linux framework supports industrial Cortex-A8 SoCs
• Real-time technology gains user conference
• Embedded Linux distro claims one-second boot-up
• Ten years after, MontaVista's Jim Ready looks forward