Intel is maintaining a four-year refresh cycle for servers in data centers as it looks to save close to US$250 million in data center costs over an eight-year period, a company executive said on Tuesday.
The company has already cut the number of data centers by half and is further looking to consolidate servers, said Diane Bryant, Intel's chief information officer, at an event on Tuesday. Intel had 147 data centers at its peak, with the now reduced to around 70.
Intel hopes to save $250 million between 2007 to 2015 by cutting costs associated with data centers, including cooling, system maintenance and support. A four-year refresh cycle for servers, which started in 2007, is already helping the company reduce such expenditure, Bryant said.
The company saved $45 million in 2008 in data center costs, but there has been a lot more scrutiny on IT expenditure this year, Bryant said. Intel decided a four-year refresh cycle for servers would be optimal as older servers eat up financial resources and cost more to replace. Intel hopes to cut data center costs by implementing faster chips, consolidating servers and putting more applications in virtualized environments, Bryant said.
Intel has consolidated servers by replacing 10 single-core Xeon chips with one Nehalem-based quad-core Xeon chip. That has helped reduce the hardware in data centers while increasing overall server performance, Bryant said. The company also cut hardware acquisition costs and related overhead costs per server, like energy and maintenance.
A big chunk of data-center expenditure involved cooling servers, Bryant said. Implementing more power-efficient servers has helped reduce energy costs, but Intel has struggled in identifying an "efficient data center." Cooling costs relate to the power efficiency of servers, a metric that has been hard to calculate, she said.
Intel is working with U.S. government agencies like the U.S. Environmental Protection Agency to measure power efficiency in different server states from idle to maximum usage, Bryant said. The EPA issued Energy Star ratings for servers in May, with the main metric criteria being the efficiency of a server's power supply and the power consumed at idle.
The company is also using technologies to squeeze out maximum server performance by maintaining high utilization rates. Intel has about 100,000 total servers, of which 80,000 are in the high-performance computing environment. Intel looks for an 85 percent utilization rate in the HPC environment without overloading the servers, Bryant said. The company has 20,000 "office" servers for normal tasks, where the company maintains a 65 percent utilization rate for maximum efficiency.
Getting applications out of dedicated hardware and into virtualized environments is one way Intel manages to attain high utilization rates, Bryant said. At the same time, Intel wants to make sure it reaches the utilization threshold without overburdening systems.
"Back two or three years ago, when virtualization became the focus, when everybody's data centers were running at 5, 10 or 15 percent utilization, the focus ... was to drive up utilization levels through consolidation and virtualization," Bryant said.
Intel is already experimenting in numerous ways to cut energy costs. Last year proposed it experimented with a data center that uses minimal air conditioning. It is also working with academia and companies like Hewlett-Packard and IBM to determine the best techniques to cool equipment in data centers.