Which Company Will Define the Cloud Decade?
April 25, 2012
In the history of computing, an innovation has defined each decade and its technology leadership.
- The 1950s were about mainframes and IBM(IBM).
- The 1960s were about minicomputers and companies such as Digital Equipment Corp.
- The 1970s were about the microchip and companies such as Intel(INTC).
- The 1980s were about local networks and companies such as Novell(NOVL).
- The 1990s were about the Internet and, to a large extent, Microsoft(MSFT).
- The 2000s were about gadgets and, of course, Apple(AAPL).
We’re now into the third year of the 2010s, and it’s clear that the computing theme of our time is cloud computing. The question is: Where will the leadership lie? Who will be the prime beneficiary?…
Some of the biggest price-to-earnings ratios in the land are to be found at cloud technology outfits VMWare(VMW), Rackspace(RAX) and Red Hat(RHT).
The idea is that the software that people choose to run their clouds — the "cloud stack" — will grow these companies, as every enterprise works to replace client-server computing with more cost-efficient clouds.
Maybe. But a key to cloud development is the favoring of open-source software, where the software’s benefits benefits flow mainly to users, not vendors.
VMWare could be the cloud’s Apple or its Microsoft, because its vSphere infrastructure is largely proprietary. Or, if the money is to be found in cloud-hosting and open-source, Rackspace could win the prize. If open-source itself is the key, maybe Red Hat will triumph.
As you can probably tell, I’m skeptical about those possibilities. Not that the companies are bad investments. They just may not define cloud when the history of our time is written.
Another way of looking at this is in terms of services. Software as a Service is the result of using clouds, delivering enormous bang for the customer’s buck. That’s what Salesforce.com(CRM)is claiming. That’s really what Amazon.com(AMZN) is about, with the retailing giant used to generate demands for digital files and cloud infrastructure. I own a little Amazon, but when you put its earnings next to its price, it’s nearly fully valued.
The excitement about Facebook is really about cloud. Facebook is, at its heart, a cloud application. It could not have scaled as quickly as it did without cloud architecture. It’s a wonderful demonstration of the power cloud technology has.
But I think it’s more likely that the people hyping Facebook to you right now are selling stock. The "social computing" boom feels to me like the dot-com bubble, and in many ways it has popped, with the shares of many social computing outfits trading at less than their offering prices. My view is: Let the private investors take that hit. Wait until there is some "there" there with Facebook. Let’s see some operating numbers.
So where does cloud money go? I think it should go first to the company that first built a cloud and that has understood cloud economics longer than anyone.
I think it should go to Google(GOOG). (I’ve got a little taste.)
My view is that cloud comes down to cost. Yes, cloud technology, by itself, saves money. No more single-server or "per-seat" licenses. But once everyone has a cloud, a cloud shakeout is inevitable. And Google has had a ruthless focus on cloud costs since it started — that’s its real secret sauce.
Google pioneered using low-cost PCs for its cloud. It bought its own dark fiber a decade ago. It took versions of itself — mini-Googles — and sent them out on trucks to be plugged in so it could save money on moving data cross-country. Its efforts in energy are all about cost, including opening the windows on data centers so it won’t have to cool its computers with expensive air-conditioning.
If Google keeps its eye on this ball, if it keeps focusing on keeping its cost per million transactions lower than those of any rival, its success can still define this decade. And the current P/E is about 18. It’s a bargain.