The self-created risk in Broadcom's big VMware kiss-off

Arise, the new Migration Nation

Opinion It looked like the most cynical corporate move of the last couple of years, at least in IT. A year and a half after it was announced, so it has proved to be. Broadcom's $61 billion buyout of VMware was expected to follow a familiar formula – price hikes, product abandonment, complete ecosystem devastation – and the creation of two classes of people.

The first class? Rich people who got a lot richer. The second? Anyone not aligned with the vision of "large enterprise only" mantra, who may now leave the New Protectorate of VMware by Broadcom, please. Like now. Citizenship revoked. Bye.

If you've never been in or dependent on a company violently acquired and stripped to its bones by sharks, you may not recognize the screams of betrayal and anger from those whose careers and clients just became so much capitalist chum. These emotions are fully justified. Remember them: Broadcom's master stroke may not play out as it hopes, with those it has abandoned as the cause.

With virtualization, it's not the product that matters as much as the underlying technology, and Broadcom can't buy that. Hardware virtualization is one of the most recent core CPU technologies to make the mainstream with Intel's Pentium 4 in 2005 and Arm with the Cortex-A7 in 2011. VMware itself launched in 1999. You don't need explicit hardware support to do virtualization, in the same way you don't need a dentist to pull a tooth. It's just a heck of a lot faster, less painful, and more reliable.

Hardware seemed magical at the time, turning complete computers into a mere software image that actually worked as you'd expect. For experimenters, enthusiasts, and developers, it was a whole new dimension in exploring options, multi platform environments, and keenly following the retail price of RAM. For enterprises of all sizes, it was catnip for efficient deployment, use of resources, and managing fleets of seats. For the cloud, it was the very spark of life.

All these things remain true enough today for Broadcom to drop a figure close to the annual GDP of Uruguay on VMware's share of the market (if you don't find that statistic obscene, try harder).

But what makes virtualization unique is also an Achilles' heel made of Kryptonite, if you're hell-bent on creating an unbreachable enclave around a corporate gold mine.

Virtualization's entire purpose is to vanish entirely by precisely mimicking hardware. It's not like an operating system, which presents a complex, unique set of APIs to functionality. It's not like application software that relies on a unique user interface, use of data, and interoperability with other components. Shifting either aspect of an enterprise's IT means uprooting the experience and workflow of everyone, to say nothing of the technical joys. That's so unpalatable that avoiding it has made Microsoft worth $3 trillion (the GDP of France, since you're curious, a country with an aircraft carrier and its own frigging nukes). But if you change your virtualization system across a corporate, no matter how huge, nobody outside the engine room will be any the wiser.

Migrating virtualization platforms is, of course, full of horrors and sphincter-tightening surprises in itself. You don't want to do it unless someone has thrown you out of the balloon at gunpoint. Which, critics say, is precisely what Broadcom has done to an enormous cohort of its supporters, customers, and practical experts.

All these people have invested money, time, and often passion in one flavor of virtualization magic. It feels like betrayal and loss because that's what it is. Yet a people dispossessed can see itself as one or two things – a tribe outcast, or a nation in the making. In this case, a nation of virtualization experts who know all the details of a major technology and what makes it work. You couldn't ask for a better bedrock on which to build a migration industry with the chops and motivation to create first the proof of concept, then the fully functional roadmap for anyone who wants to follow them out of the New Protectorate.

This is already happening in the Discords and the YouTube comments and wherever the new tribe is gathered. Among the anguish and anger at fears fulfilled, the discussions are already advancing – what is the way forward, technically, to support ourselves and our clients? How are we to guide the on-prem versus cloud discussion with the CIO? What now? Others are answering with experiences good and bad, with advice and warnings, an entire industry's worth of experience and motivation in the process of solving a problem they're already very good at.

Having deliberately written off that part of their ecosystem, it's doubtful Broadcom would care. After all, it's nothing to do with the large enterprise cash cow from its point of view. Who among the little people knows how to operate in our multibillion-dollar paradise? Two answers to that: you have no idea how many big clients the little people have, and a whole load of corporate expertise just walked out your door – to say nothing of the ones you've just pushed out.

Broadcom has made a huge wager on corporate inertia, price insensitivity, and lack of viable alternatives. That's normally a fair wager, given the state of most corporations and the huge cost of entry for would-be competitors.

But the equation is different with virtualization. It's no "shall we replace Intel with AMD" but it's closer to that than "Let's ditch Windows!"

Proofs of concept will not be long in coming, and they will be coming from those who have to move, and the concepts will include the wisdom of avoiding proprietary solutions that can be weaponized. You can't close down open source, but you can scale it to the largest enterprise.

This would all be powered by the new nation of experts with skills, experience, contacts, and necessity. And if you want to add a twist of vengeance to that mix, feel free. ®

More about


Send us news

Other stories you might like