Utility computing can save SMEs £10K a year

An SME moving to utility computing, using a hosted server and thin clients, could achieve “£10K saving per year”, according to Ben Rapp, CEO of support and services company Managed Networks.

Share

An SME moving to utility computing, using a hosted server and thin clients, could achieve “£10K saving per year”, according to Ben Rapp, CEO of support and services company Managed Networks.

Speaking at the Green Hosting to Utility Computing conference in London yesterday, Rapp said his model, based on a 10-user SME, was broadly applicable to all utility computing services.

He calculated that “setting up and running 10 computers on a Windows network over four years costs £47K. Using utility computing instead brings this down to about £37K, with no capital expenditure. It is a 21% saving, and a 57% saving on 6% capital cost.”

This model also delivers a 60% carbon savings, according to Rapp. His model is based on a business buying new thin clients. The re-use of old computers as thin clients saves more cash but not as much carbon. The model assumes a cost of £0.10 per kilowatt hour for electricity, although many companies will have deals to pay less than this.

With utility computing, an organisation’s server is hosted by the company, which also runs its applications in a virtual environment. The hosting company is responsible for the software licences, server support and maintenance, and its energy consumption. Rapp describes the utility model as “renting rather than buying”.

It is the money, cap-ex, and personnel savings that attract business, not the carbon savings. “There is no evidence that anyone has ever bought ‘green’ for that reason alone.

It just so happens that it can take down their carbon footprint by 50 to 60%. Customers won’t pay more for the privilege of buying ‘green’, and they shouldn’t have to. Green is energy efficiency and you shouldn't have to pay more for that”.

The model includes the embedded carbon of devices as well operational expenditure, which presents additional problems of accuracy.

“Data for embedded carbon is hard to come by and old. Assumptions have to be made about increases in efficiency since this data was gathered.” The model assumes that about five tonnes of CO2 is used in the manufacture of 10 desktops. It estimates that about 18.2 tonnes of CO2 is used when running an SME’s network. This is reduced to 7.5 tonnes when a hosted service is used.

Promoted