Microsoft has been operating data centers since 1989. In 1989-2004, data centers were used to address early computer systems that require controlled temperature environment to operate. Between 2004-2007, they began to design their own large facilities to provide online services with large range of solutions. In 2007, 1st generation data center was opened in Washington which has 100,000 of high high performance processing storage service and high density racks that are separated by hot and cool air which are kept on traditional concrete floors. Chillers and air handling equipment ensures control over environmental conditions. To ensure continuous electricity all important arrangements has been made. High speed robust optic network connects this data center with other major hubs. The modular approach of data center has helped to reduce costs, deployment time and has increased overall efficiency.
In 2009, a new 3rd generation facility data center was opened in chicago and was first implementation of modular unit to reduce cost and deployment time. Large units were transported directly to center with the help of trucks and were installed and connected to cooling system. Water side economization was used to cool facility without requiring high electricity. A traditional database has a PUE(Power Usage Effectiveness = Total Facility Power/IT Equipment energy, which is used to measure total power consumption of data center) of 2.0 but this data center has a PUE 1..15-1.22.
In 2009, this approach was further implemented in Dublin. In this server are cooled by airside economization in which air is pulled inside to cool server. Evaporative coolers vaporizes water into air to absorb heat and air bypassing feature improve operation efficiency and maintains constant room temperature. It uses less than 1% of annual usage consumption of traditional data center. It has PUE of 1.25.
In 2010, generation 4 data center was built which significantly reduced water and power use. It utilizes air cooled IT pre-manufactured components or IT packs. Pre-assembled modular component are shipped onsite and reassembled. IT packs reduces construction time by 2 years and capital cost by 30%-50%. It also has self contained ups with air handling and evaporative cooling system. This data center maintains PUE of 1.15-1.2 and energy is provided from 100% renewable hydro power, Security is an important aspects of data center so all the measures are followed to secure the data according to priority.
Inside Data Center
In 2009, a new 3rd generation facility data center was opened in chicago and was first implementation of modular unit to reduce cost and deployment time. Large units were transported directly to center with the help of trucks and were installed and connected to cooling system. Water side economization was used to cool facility without requiring high electricity. A traditional database has a PUE(Power Usage Effectiveness = Total Facility Power/IT Equipment energy, which is used to measure total power consumption of data center) of 2.0 but this data center has a PUE 1..15-1.22.
In 2009, this approach was further implemented in Dublin. In this server are cooled by airside economization in which air is pulled inside to cool server. Evaporative coolers vaporizes water into air to absorb heat and air bypassing feature improve operation efficiency and maintains constant room temperature. It uses less than 1% of annual usage consumption of traditional data center. It has PUE of 1.25.
Aerial View of Dublin Data Center
In 2010, generation 4 data center was built which significantly reduced water and power use. It utilizes air cooled IT pre-manufactured components or IT packs. Pre-assembled modular component are shipped onsite and reassembled. IT packs reduces construction time by 2 years and capital cost by 30%-50%. It also has self contained ups with air handling and evaporative cooling system. This data center maintains PUE of 1.15-1.2 and energy is provided from 100% renewable hydro power, Security is an important aspects of data center so all the measures are followed to secure the data according to priority.
No comments:
Post a Comment