When Florian Leibert, co-founder of the hot infrastructure start-up Mesosphere, Inc. told theCUBE last week that consolidation is coming to the infrastructure market, he probably didn’t expect his predictions to come true quite so quickly.

Leibert was talking about the long-term impact of commodity components and open source software, but an impatient Wall Street is pushing the issue a little harder. Last week Oracle Corp. disappointed investors with earnings that missed expectations and a confusing earnings call that had many outlets reporting that founder Larry Ellison is stepping down (he isn’t). The Wall Street Journal is reporting today that EMC Corp. is seeking a merger partner as it confronts pressure to break itself up and deals with executive succession uncertainties. It reportedly held merger talked with Hewlett-Packard Co. a deeply troubled tech firm whose most profitable product line is ink.

Those aren’t the only onetime IT titans that are floundering. Microsoft Corp. is laying off 18,000 people. CA, Inc. is struggling to reinvent itself. Longtime enterprise stalwart Compuware Corp. is selling out to a private equity firm. Dell Corp. also went private. If you had bought a basket of stocks that included CA, Oracle, HP, Microsoft and EMC 15 years ago, your portfolio would have lost money today. Their 10-year stock charts look like an EKG.

Changing of the guard

The companies that are fueling the innovation that will drive the data center of the future are not the ones that dominated it in the past. They’re Amazon.com, Inc., Google, Facebook Inc., Twitter, Inc. Yahoo!, Inc. and a host of other born-on-the-web companies that are beginning to package their core technologies into products and spin off new companies.

Many people scratched their heads when Amazon launched Amazon Web Services in 2006, but the Internet retailer was really just an early mover in its market. Like many Internet companies, Amazon has had to solve a lot of thorny technical problems to scale its business at an unprecedented rate. The same holds true for Google, Facebook, Yahoo! and others. No one had ever processed 40,000 queries per second before, so Google had to figure out how to do it. Technologies like Hadoop, Kubernetes, Docker and Mesos, which are set to transform corporate IT, were built because their authors essentially had no choice.

In contrast, traditional born-in-the-data-center companies had relatively simply challenges. Their customers had closed, vertically integrated environments with captive user bases. If a query took 30 seconds to run, the user had no choice. But to the Googles and Amazons of the world, the consequence of a three-second delay is bankruptcy.

The reason we haven’t seen born-on-the-web companies enter the enterprise infrastructure market more aggressively until now is because they haven’t had the time. They were too busy scaling their businesses. Now that Internet growth is moderating, they’re turning their attention to new revenue opportunities, and Big Data and the corporate data center are the most obvious choices.

Traditional vendors have been caught flat-footed by infrastructure innovations that are set to sweep the industry. They didn’t expect competition from search engine and ecommerce companies, and they’re scrambling to respond. But it’s already too late.

Those big companies have a lot of cash, and they can continue to grow through acquisition, but owning a lot of legacy software is a drag on innovation. Ask CA and Compuware about that. The smartest traditional software vendors should be scouring the ranks of infrastructure startups to find the next VMware. Merging with other industry megaliths isn’t the answer. A merger of EMC and HP doesn’t make sense. Two wrongs don’t make a right.

photo credit: Bill Gracey via photopin cc