Telehouse Green: Innovation Through Collaboration
December 1, 2016
How the Open Compute Project Is Transforming Data Center Infrastructure and Hardware
As Albert Einstein once stated, “Problems cannot be solved by the same level of thinking that created them.”
In the data center and colocation industry, where copious amounts of energy used to power critical infrastructure cause significant strain on natural resources and the bottom line of facility owners and operators, the need for a new level of thinking has become an existential requirement. To meet this challenge, the data center community has been forced to shift its longstanding and entrenched perspective on hardware and infrastructure to become more dynamic, inventive and holistic in its approach to design.
Enter the Open Compute Project, which was inspired by the creativity and collaboration exemplified by open source software. The Open Compute Project officially launched in 2011 when Facebook decided to share its design for the world’s most energy-efficient data center with the public. Soon after, Intel®, Rackspace, Goldman Sachs and Andy Bechtolsheim, the electrical engineer who co-founded Sun Microsystems and later became an early investor in Google, enlisted their support.
The mission of the Open Compute Project is based on a simple, yet powerful concept. Members of this community believe that openly sharing ideas, specifications and intellectual property is the key to maximizing innovation and reducing complexity in the tech components needed to support the growing demands on compute infrastructure. Today, with hundreds of participants actively collaborating, the Open Compute Project is transforming data center infrastructure and hardware design with a focus on energy and material efficiencies.
The Data Center: A Single, Ubiquitous Ecosystem
While traditional data center design often occurs in isolated components such as the building, servers and software, by contrast, the Open Compute Project evaluates the collective influence of all components within the data center environment. This unique approach to viewing the data center as a single, ubiquitous ecosystem leads to optimized energy and material use, as well as reduced environmental impact. Three core aspects of the Open Compute Project’s approach to data center infrastructure and hardware include enhanced rack design, localized back-up power and evolved machinery.
Open Compute Project server racks are created with minimalist design and more efficient use of available space. By removing traditional rack facades and bezels, this construction minimizes the amount of metal surrounding servers, allowing air to flow freely and keeping the rack naturally cooler to reduce energy usage. In addition, new designs enable racks to hold more servers, thereby benefitting from power density. This approach strips out the excess components of a rack, creating up to 50 percent more available server space.
Traditionally, data center operators run two unique power supplies to the server to ensure continuous uptime in the event of a power failure. While this approach to power backups may be effective, it is an extremely wasteful practice. To combat this, Open Compute Project designs share power supply for the entire rack and install batteries directly into the structure, eliminating the need for large Uninterruptible Power Supplies (UPS).
Finally, there is the Open Compute Project’s promotion of evolved machinery. Traditionally, legacy vendors built hardware that was perceived as so delicate that servers were run far below their optimal effectiveness. The result? More machines were used than necessary, energy was wasted, servers run at sub-optimal rates would give off additional heat, and even more power was needed to keep them cool.
The Open Compute Project promotes commodity hardware rather than buying gear from legacy vendors. This evolved machinery gives power back to facility operators by transferring the intelligence out of the machines and into the software, creating resilient computational systems while encouraging a mindset of self-service and optimal usage.
From high-end, mission-critical servers to customized colocation equipment, Telehouse provides comprehensive, customized and scalable hardware platforms that save your company capital expenditures while realizing predictable operational costs.
