FOR IMMEDIATE RELEASE

SiCortex: New ENERGY STAR® Ratings for Computer Servers a Step in the Right Direction

Future Energy Star specifications for large multiprocessor systems will further drive energy efficiency, reducing data center energy consumption
May 17, 2009

Maynard, Mass. — May 18, 2009 — ENERGY STAR® recently announced a new ratings system for low-end computer servers. These specifications mark an industry milestone that will bring energy issues to the forefront in informing purchase decisions of computer servers comprising up to four processors. Impacting the small business and departmental server markets for both buyers and manufacturers, the ratings are a “win” for multiple energy stakeholders: consumers, manufacturers who build-in energy efficiencies, economies and the environment. “Setting standards for energy consumption is clearly an important step in the right direction – tackling head-on a product category that consumes vast amounts of electricity,” stated Christopher Stone, president and CEO of SiCortex.

According to industry reports, at the current rate of energy consumption growth, the carbon footprint of data centers in the US will eclipse the airline industry by 2020. “The next tier of specifications will have an even greater impact, targeting the large, multiprocessor computer systems, which are perhaps the biggest culprit in data center energy consumption,” added Stone. “We at SiCortex applaud ENERGY STAR’s efforts in taking on this important challenge.”

SiCortex has been actively involved in this issue, having recently proposed the Green Performance Computing Index ( GCPI). The GCPI is an industry benchmark-based metric to rate the relative energy efficiency of the leading high-productivity computing (HPC) systems manufactured by Cray, Dell, HP, IBM, SiCortex and others. The index compares the results from a suite of 18 benchmarks measuring different aspects of computer performance in terms of energy efficiency, and provides a weighted average result. For those users whose applications rely on a less balanced system, the GCPI now features an online calculator that allows users to set their own weightings and compare systems accordingly. Pending the availability of the Tier 2 ENERGY STAR specifications for large multiprocessor systems, the GCPI provides an objective tool to guide purchase decisions for these systems.

There is significant need for tools such as the GCPI and ENERGY STAR ratings. Enterprises and research facilities alike tend to replace computers in three year cycles, and energy efficiency improvements can help reverse the data center electricity crisis over time. McKinsey projects that within the next 30 months, 90 percent of companies running large data centers will need more power and cooling, but most will reached the limits of their energy capacity. Green computers, systems that require less electricity than their carbon slurping counterparts installed in the nation’s data centers today, might be the best bet for data center managers facing this problem.

According to information released by ENERGY STAR, the ENERGY STAR Tier 1 server ratings are based on the SPECPower ssj_2008™ benchmark and apply to servers with up to only four processors. High-productivity computers, network and storage equipment, server appliances and blade systems will be ineligible under this rating system until ENERGY STAR Tier 2 is launched.

While SiCortex endorses the Tier 1 ratings, the company is looking ahead to the Tier 2 benchmarks, which will ultimately be more relevant to its core customers. According to Stone, servers with more than four cores dominate the scientific research world, and most of the data centers associated with academic institutions and research laboratories in the U.S. run high-productivity computers.

“Presently, the evaluation of energy consumption is a fundamental consideration in the acquisition of scalable server technologies,” said Addison Snell, General Manager of Tabor Research. “Common benchmarking tools like ENERGY STAR and GCPI will not only make it easier for buyers to determine the power and cooling tradeoffs between platforms, but also help the industry as a whole by enabling data center managers to effectively maximize productivity within their facilities budgets.”