Utility Computing Finds its Sea Legs

From Innovations, a website published by Ziff-Davis Enterprise from mid-2006 to mid-2009. Reprinted by permission.

Nearly a decade ago, I worked at an Internet startup in the crazy days of the early Web.  Success demanded speed, rapid growth and scalability.  We were frequently frustrated in those days by the demands of building out our computing infrastructure.

The company burned more than $1 million in that pursuit. Racks of Unix servers had to be acquired and configured.  The equipment was set up at a co-location facility that required on-site care and feeding by technical staff.  Installation and testing took weeks.  Maintenance was a time-consuming burden requiring a staff of technicians who had to be available at all hours.  At least twice during the first year, server crashes took the company off-line for more than a day.  Stress and burnout were a constant issue.

Today, I suspect the company would do things quite differently.  Instead of acquiring computers, it would buy processing power and storage from an online service.  Startup times would be days or weeks instead of months.  Scaling the infrastructure would require simply buying more computer cycles. There would be no cost for support personnel. Costs would be expensed instead of capitalized.  More importantly, the company would be up and running in a fraction of the time that was once required.

The current poster child of utility computing is, of all companies, Amazon.com.  An article in last week’s Wired magazine describes the phenomenal success of the initiatives called S3 and EC2 that Amazon originally undertook to make a few dollars off of excess computing capacity. Today, the two services count 400,000 customers and have become a model that could revolutionize business innovation.

That’s right, business innovation. That’s because the main beneficiaries of utility computing are turning out to be startups. They’re using the services to cut the time and expense of bringing their ideas to market and, in the process, propelling innovation.

The utility computing concept has been around for years, but questions have persisted about who would use it. Big companies are reluctant to move their data offsite and lose control of their hardware assets.  They may have liked utility computing in concept, but the execution wasn’t worth the effort

It turns out the sweet spot is startup firms. Many business ideas never get off the ground because entrepreneurs can’t raise the $100,000 or more needed for capital investment in computers. In contract, Amazon says it will transfer five terabytes of data from a 400G-byte data store at a monthly fee of less than $1,400. If you use less, you pay less. It’s no wonder cash-strapped companies find this concept so appealing. Wired notes that one startup that uses Amazon services dubbed one of its presentations “Using S3 to Avoid VC [venture capital].”

Now that companies are getting hip to this idea, expect prices to come down even further. Sun already leases space on its grid networkIBM has an on-location variant. Hewlett-Packard has an array of offerings. There are even rumors that Google will get into the market with a free offering supported by advertising. And, of course, there will be startups.

The availability of cheap, reliable and easy-to-deploy computing services could enable a whole new class of entrepreneurs to get their ideas off the ground.  It’s just one more example of IT’s potential for dramatic business change.

Leave a Reply

Your email address will not be published.