The Linux Foundation published a study recently that pegged the value of the Linux kernel at $1.4 billion. Another $10.8 billion would be required if one were to develop the entire stack of components that makes up the platform from scratch. I won't be delving into the details here, though Ars Technica gave a pretty good round-up of the methodology, should you want to find out more. The whole concept of trying to figure out how much you'd have to pay to develop what has been built in an open source model is fascinating.
The study adds another layer to the ways that SMBs benefit from Linux.
As I mentioned in my earlier post on the release of OpenOffice 3.0, SMBs are often disadvantaged when it comes to acquiring software licenses in that they are tend to be too small to qualify for special volume discounts, compared to their to mid- or enterprise-sized counterparts. Ditto for operating system licenses.
On my end, I have personally deployed Linux-based firewalls, e-mail, Web, application and database systems when working in a number of SMBs. For a period of time, I maintained a high-availability Linux-based server used to automate a factory machine. Attempting to get the above-mentioned services implemented in Windows - even if it were possible - would have resulted in a substantial capital outlay. As such, the use of Linux machines can lead to some big cost savings.
The stability of Linux can be a question for SMBs that don't have experience with it, but the fact that many companies have built products on top of the Linux kernel, including all-in-one appliances like the SnapGear SG565 and hardware gadgets such as the Yoggie Pico hardware firewall serves as a cue to newcomers that it's not likely to be an issue.
In certain circumstances, considering Linux as an alternative platform for server-side tasks makes a lot of sense.