A few years ago I was honored to keynote an ApacheCon in Atlanta, and the prize was interviewing Doug Cutting, incoming chairman of the Apache Software Foundation, which organizes and boosts open-source software. Open-source software is freely available. It can be fixed or upgraded by anyone who downloads it.
Cutting had previously created Hadoop, named for his son's stuffed elephant. Hadoop breaks down big math problems into a collection of smaller problems, then passes them around and collates the answers. Cutting told me it's like a news editor handing out and collecting assignments. Such "big data" solutions are at the heart of what clouds do well.
When we talked, in the fall of 2010, there was still a question about whether open source could make money, and whether big companies such as Intel would ever understand that software needs to be free in order to grow.
By paying an estimated $740 million to $760 million for 18% of Cloudera, much of which is going to existing shareholders, Intel has answered that question in the affirmative.
The deal offers more evidence, as though any were needed, that big tech companies see open source as the only way to stay relevant in cloud computing.
Hewlett-Packard (HPQ) and IBM (IBM) have both committed to open-source OpenStack for their cloud infrastructure. VMware (VMW) recently placed the code for its Cloud Foundry platform into an open-source foundation.
The problem is that, when it comes to cloud, big tech remains at the bottom of the profit stack.
Cloud is an architecture, not a product the way a mainframe is a product or a PC is a product. You can't take a cloud and ship it to someone on a truck, then plug it in. Intel x86 chips are used in cloud data centers, but such centers are built with cheap commodity chips, not the high-end server chips on which fat margins are built.
Turning cloud into a high-profit product requires more than a big box and a cloud operating system. It requires a box that is cost-competitive with do-it-yourself solutions like Google (GOOG), Amazon.com (AMZN) and Facebook (FB) use. And it requires an integrated software stack -- an infrastructure, platform and applications -- that justifies a premium price.
Hadoop could be part of that solution, and Intel's gain in Tuesday's trading reflected optimism that it would.
By integrating a big data solution and a full cloud stack, by supporting that software as part of the larger buy, Intel may be able to put a cloud in a box and ship it profitably, with back-end support revenue attached.
To some this appears a bit desperate, but these are desperate times for big tech companies. Amazon dominates in cloud infrastructure, Google has the largest clouds, Facebook is growing fast, Microsoft (MSFT) is a software company, and none has the slightest need for the high end of what big tech is selling.
Intel's Cloudera investment may start to change that, or it could just turn out to be a profitable investment, as its 2007 buy of part of VMware was a profitable investment. But it's just the start of a solution, not the solution itself. In the race to the clouds, big tech remains far behind.
That's the real message of this story.
At the time of publication the author owned shares of AMZN and GOOG.
This article represents the opinion of a contributor and not necessarily that of TheStreet or its editorial staff.