IBM 'Gas Gauge' Monitors Mainframe Energy Usage

IBM has added a mainframe gas gauge to more than 1,000 z9s deployed since May 11 to monitor energy usage and cooling statistics.

InformationWeek Staff, Contributor

October 11, 2007

3 Min Read
InformationWeek logo in a gray background | InformationWeek

The Environmental Protection Agency in August asked server manufacturers to develop "miles per gallon" ratings for their equipment that would provide users with accurate assessments of energy efficiency. IBM on Thursday announced that it would begin providing typical usage ratings for its line of z9 mainframe computers.

The usage ratings were developed using data collected from more than 1,000 z9s deployed to customers since May 11 when IBM added a "mainframe gas gauge" feature to the computers that monitors energy usage and cooling statistics. As a result, IBM has found that typical energy use by the systems is normally 60% less than the maximum ratings.

IBM will provide typical energy use numbers by model for its z9 mainframes, and will continue to monitor systems in the field to adjust the ratings on a monthly basis, said Dave Anderson, an IBM green consultant, in an interview.

"Over time every vendor is going to be asked to provide typical energy use numbers for their equipment," Anderson said. "It's what the EPA wants, and this allows us to move beyond simple performance benchmarking to energy benchmarking."

The mainframe gas gauge also can help customers monitor and manage their own energy use, and provide accurate documentation that can be used for planning for more efficient processes and to present to utility providers for potential rebates or rate reductions, he said. Having accurate typical use numbers also can ensure that a business doesn't over-provision for its power and cooling needs.

The gas gauge monitors energy and cooling statistics using internal sensors built into the mainframes. The statistics are collected within the System Activity Display, which can correlate the energy consumed with work performed. A Power Estimator Tool also is available to aid energy efficiency planning efforts. It calculates how changes in the system configurations and workloads can affect the total energy "envelope," including the power needed to run and cool the machines, Anderson said.

The August EPA report found that data centers in the United States are taking up an increasing amount of energy. The report found that data centers use about 1.5% of all electricity consumed in the country, and that total use had doubled in the past five years. The EPA warned that usage will double again if current data center power consumption trends continue.

But Anderson points out that while data centers have increasingly become villains in the energy efficiency battle, the operations also are responsible for enabling a reduction in some energy requirements by improving the efficiency of businesses operations and enabling efforts like telecommuting.

"We can now have some accurate numbers that reflect measured energy use," Anderson said. "Business can approach their energy providers with some hard facts about their consumption and overall efficiency efforts."

Thursday's announcement is part of the Project Big Green initiative IBM announced in May. The company has said it will spend $1 billion this year in increasing the efficiency of IBM products.

IBM announced in August that as part of Project Big Green it was working to consolidate about 3,900 mostly x86-based servers onto about 30 mainframes running the Linux operating system. The company anticipates the new environment will cut energy consumption by about 80%, and save IBM more than $250 million over five years in energy, software, and system support costs.

IBM believes more businesses will look to consolidate operations on mainframes, and believes providing accurate power and cooling usage numbers will help demonstrate the efficiency of using a mainframe installation, Anderson said.

Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.

You May Also Like


More Insights