This audio was created using Microsoft Azure Speech Services
These terms have been discussed and promoted ad nauseam for years. Perhaps so much so, that my mere use of them in this title risks significantly reducing the number of people who read this far. Regardless, there’s no denying the conceptual benefits of industrialized modularity. And, in most cases, the data center physical infrastructure industry is quickly evolving its components, solutions, and services to be increasingly so. Indeed, for many people in this industry, the on-going evolution towards standardized modularity which enables faster build-outs and scaling is inevitable.
This evolution is far from being mature or fully baked, however. Data center design managers, planners, operators and managers are left facing a number of confusing, poorly defined terms describing modularity such as pods, containers, zones, clusters, rooms, rows, busses, etc. The definition of modularity itself remains vague and is further complicated by the fact that it can be (and is) applied at multiple levels from a single device, such as a UPS, to an entire data center. Further, the benefits of these concepts have sometimes been exaggerated and touted in areas where it’s not necessarily appropriate.
As a result, people responsible for designing and implementing a data center project may find it difficult to actually specify modularity in relation to the physical infrastructure. At what point is a data center considered “modular”? What is the appropriate “chunk” size for a given module? What are the deployment rules for that module and how is it related to modules for other subsystems (e.g., for a unit of cooling capacity, how many racks are supported?) Questions like those needs to be answered in order to properly specify modularity for a project or new build.
The founder of the Schneider Electric Data Center Science Center, former CTO, and now the Chief Innovation Officer for Schneider Electric’s IT Business, Neil Rasmussen, has recently authored a new white paper that provides a sound framework for how to do this. White Paper 160, “Specification of Modular Data Center Architecture”, defines what is meant by modularity and defines the terms used for describing and specifying modularity in relation to the physical infrastructure of data centers including space, power and cooling. A graphical method for describing a modular architecture is presented. The feasibility of standardizing and industrializing modularity is examined. He also shows how data center modularity can be effectively applied and specified, and how the approach should vary with the application. Check it out…it’s free and you don’t even have to register.
