Facebook’s Open Compute Project has brought some key innovations to cloud hardware design. For example, organizations seeking to scale operations have benefited from new guidance about managing environmental conditions and supporting bandwidth-intensive services such as virtual desktop infrastructure
However, the most important facet of Open Compute Project may be the freedom it gives companies in how they procure and deploy hardware. Technology blogger Brian Madden highlighted how organizations can design and/or purchase appliances from OEMs, gaining access to systems that imitate the hyperscale operations of vendors such as Facebook, Amazon and Google. These setups often use reconfigured basic components such as CPUs, SSD chips and Ethernet ports rather than commercially integrated solutions.
“Of course we don’t actually know what’s going on inside those [hyperscale operators’] data centers today, but we can be sure those cloud providers are thinking more along these lines rather than sending RFQs to Dell and HP,” wrote Madden. “And the Open Compute Project means that even smaller DaaS and IaaS providers who don’t have electrical engineers on staff can still buy these types of systems from white box builders in Asia.”
Madden pointed out that these open hardware configurations have enabled hyperscale operators to offer DaaS, IaaS and VDI at a scale and price that organizations would be hard-pressed to match with traditional rack server configurations. On top of that, Open Compute Project designs allow companies to save space and energy in the data center, since appliances are downsized and sometimes cooled with ventilation rather than air conditioning.
In an interview with Consulting-Specifying Engineer, Jacobs Engineering design principal Kevin Dickens urged cloud storage system architects to learn from Open Compute Project. More specifically, they should move beyond legacy architectures and carefully calculate risk and environmental conditions when choosing a data center location.
