Sensing the Cloud: A Materialist Spatial Analysis of Data Centers and Critical Conceptualization

Abstract

With the ever increasing digitization of the global economy and everyday life, it is essential to understand the distribution and impacts of networked infrastructure, particularly Data Centers. While Data Centers as infrastructural objects have existed in some form or another since the advent of the internet, over the course of recent decades there has been a pronounced shift in the number of data centers being constructed, a marked increase in building size, and resource use, and dramatic changes in the types and purpose of data centers, particularly pronounced since the advent of Cloud Computing and the growth of the Public Cloud. This thesis provides an initial investigation of data centers in the United States and consists of two interrelated parts. I first conduct an exploratory spatial analysis to examine multiple aspects of the data centers regarding their spatial distribution, position within the built environment, and land-use trends. Analysis revealed the highly clustered nature of data centers within metro areas associated with other physical infrastructures such as highways and land-use change agents. Second, with the results drawn from the spatial analysis, I conceptualize data centers as materially and spatially relevant nodes to comprehend the socio-technical, political, and economic implications of data center infrastructures. I root this infrastructure solely within the means of capitalist production and accumulation.