April 26, 2012
Energy consumption accounts for a large portion of datacenter operating budgets. The added environmental impact compounded with capital costs has motivated operators to increase the efficiency and lower the cost of computing. Last year, Facebook spun up the Open Compute Project, with the goal of improving datacenter designs to meet those goals.
So far, the project has produced a number of components including a server chassis, a battery cabinet and two x86 motherboards. All of the designs are freely available to download from their website.
To rate efficiency, datacenters calculate their power usage effectiveness (PUE) ratio. The rating compares how much incoming power to energy used for computational processes. One estimate puts the average datacenter PUE at 1.8. Larger Internet players have used some unique solutions to bring that number closer to the 1.0 mark, with Google currently averaging a 1.14 PUE.
Prineville, Oregon is home to one of the world’s most efficient datacenters. Built by Facebook, the center utilizes designs from the Open Compute Project as well as innovative LED lighting and gray water facilities to receive a 1.07 PUE rating at full load.
Ken Pratchett, Manager of the Prineville data center, discussed the cost and efficiency of the Open Compute Servers, “These machines are 38 percent more efficient than another machine that you could find on the open market,” he said. “In fact, they cost 24 percent less to create.”
The combination of green technologies implemented at the Oregon datacenter earned it a Gold Certification from the US Green Building Council’s Leadership in Energy and Environmental Design (LEED). Compared to other datacenters built to code, it consumed 52 percent less energy and 72 percent less water for occupant use. It also recycles captured water for landscape irrigation.
Next month, the Open Compute Project will hold a summit in San Antonio. The two-day event will include the following workshops:
The consumer-driven demand for computer technology has spurred the need for more and larger datacenters, resulting in increased energy consumption and higher operational costs. Examples like the one in Prineville not only lead to innovation in the field of green datacenter technology, but also provide a framework for cost-effective operations as well.
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?