What Are Data Centers?

Data centers are specialized facilities that house and operate the computing infrastructure required to store, process, and distribute data at scale. They provide the controlled physical environment, reliable power supply, and high-capacity network connectivity that servers, storage systems, and networking equipment require for continuous operation. Data centers range from compact server rooms supporting a single organization to hyperscale campuses operated by cloud providers that span millions of square meters and consume electricity comparable to small cities.

As the foundational physical layer of the internet and enterprise computing, data centers are the sites where cloud services, streaming platforms, financial transaction systems, and scientific workloads execute. Their design, operation, and geographic distribution are studied across disciplines including electrical engineering, computer science, facilities management, and environmental science.

Types and Ownership Models

Data centers are classified by ownership and service model. Enterprise data centers are built and managed by a single organization for its own exclusive use, giving the operator full control over hardware choices, security policies, and configuration. Colocation facilities, sometimes called carrier hotels, rent floor space, power, and network connectivity to multiple tenants who bring their own equipment; the colocation operator manages shared building infrastructure while each tenant controls its servers. Managed services providers own both the facility and the equipment within it, offering computing capacity on a contractual basis without requiring customers to manage hardware. Hyperscale data centers, operated by companies such as Amazon Web Services, Google, and Microsoft Azure, host multi-tenant public cloud services at a scale that justifies custom server designs, proprietary networking, and direct procurement of renewable energy. IBM's overview of data center types describes how these ownership categories differ in cost structure, flexibility, and operational responsibility allocation.

Environmental Footprint and Sustainability

Data centers collectively consume approximately 1 to 2 percent of global electricity, a share that has grown with the adoption of cloud services and the expansion of AI training workloads. Energy efficiency is measured primarily through power usage effectiveness (PUE), the ratio of total facility power to IT equipment power. Well-operated hyperscale facilities achieve PUE values below 1.15, while older enterprise data centers often operate above 1.5. Water usage effectiveness (WUE) tracks the water consumed for evaporative cooling per unit of IT energy, an increasingly important metric as data centers are sited in regions with water scarcity. Major operators have made public commitments to run on 100 percent renewable energy, using power purchase agreements (PPAs) and on-site generation, though the timing of renewable supply and demand rarely aligns perfectly. AWS's data center sustainability documentation describes how geographic distribution of facilities across regions with different grid carbon intensities is used to optimize the emissions profile of cloud workloads.

Physical Location and Site Selection

The location of a data center affects its operating cost, resilience, latency to users, and regulatory environment. Proximity to low-cost or renewable electricity sources is the primary driver for hyperscale siting, leading to concentrations in the US Pacific Northwest (hydropower), Scandinavia (hydropower and cold climate cooling), and the US Southwest (solar). Connectivity to fiber networks, availability of water for cooling, seismic and flood risk, and local tax incentives are secondary factors. Edge data centers, smaller facilities placed close to population centers or industrial sites, trade some cost efficiency for reduced network latency. Data residency regulations in jurisdictions including the European Union require that certain categories of personal data be stored and processed within defined geographic boundaries, directly influencing where data centers are built. Cisco's data center overview describes how network connectivity requirements interact with site selection decisions.

Applications

Data centers support a broad range of computing-dependent activities, including:

• Cloud infrastructure hosting for enterprise applications and SaaS platforms
• Content delivery and video streaming at global scale
• Financial services: high-frequency trading, payment processing, and core banking
• Scientific and academic high-performance computing
• Telecommunications: routing voice, video, and data traffic across carrier networks