According McKinsey research paper ‘Data Centers: How to Cut Carbon Emissions and Costs,’ the world’s 44 million servers consumed .5% of all electricity in 2008, with data center emissions approaching those of small countries and projected to quadruple by 2020.
The same report estimated that 25% of the average IT budget is spent on computing resources, including facilities, storage devices, servers and staffing and that the average corporate data center is only about 5% efficient. Yikes.
Much of the energy waste in data or shared computing centers results from inefficient cooling or outdated servers. However, a significant amount of energy waste results from a failure to understand and accurately estimate the cost of data in terms of computing resources.
- Rising demand for data – Companies want real-time access to complex analytics and historical data to support business decisions. To support this requirement, multiple copies of the same information are often duplicated in multiple systems.
- Decentralized decision making – Individual business users make data usage decisions without considering the impact on total capacity or cost, such as a sales manager who wants real-time access to sales pipeline analysis or an employee who doesn’t delete old emails.
- Excess capacity – Companies typically purchase excess capacity to accommodate extreme usage scenarios or future expansion, rather than implementing more efficient computing techniques such as virtualization or fine tuning software to minimize server use.
- Lack of financial accountability – Total computing costs are tracked as an expense but not energy utilization or waste.
Until I started researching corporate sustainability I never considered the carbon footprint of data – and then I felt like a schmuck for all my undeleted emails. But although companies and individuals can try to make more responsible choices about how they save and use data, the demand for data isn’t going to decrease.
That makes it imperative to dramatically increase computing efficiency, which is how Software as a Service (SaaS) solutions are already making a difference. SaaS is a software delivery model that allows companies to share computing resources rather than building and maintaining their own. So, instead of installing business software on your own premises the software is hosted in a shared data center.
What’s so green about that? Because shared computing is part of their core business, SaaS vendors have a strong incentive to build state-of-the-art data centers. And each company that moves to a centralized computing model is able to decommission their own servers and – probably inefficient – data centers.
As a result, total global CO2 emissions also decrease significantly as more companies elect to share computing resources in modern, efficient data centers.
There’s a lot to be said for SaaS business solutions, like the fact that they don’t require upgrades. Anyone who’s experienced a software upgrade will appreciate this.
But they also help reduce the carbon footprint of data.