Micro data centers, which have a small footprint, deliver the capabilities of traditional data centers in a more compact form and are often used for edge computing applications. According to Gartner, it is projected that by 2025, 75% of data generated by enterprises will be processed outside of the cloud or centralized data centers, with a significant portion being handled through micro data centers.

Micro data centers are compact, self-contained computing facilities that include components like servers, storage, networking, power, and cooling on a small scale. These portable units can be deployed indoors or outdoors, support IT loads of 1-100 kW and occupy less than 100 square feet of space.

Micro data centers are transforming the way organizations process and store data, offering a compact and efficient solution for various industries. In this article, Dgtl Infra explores the key components, advantages, and use cases of micro data centers, shedding light on their growing importance in digital infrastructure. From edge computing applications to 5G networks and the Internet of Things (IoT), discover how micro data centers are making a significant impact and explore the leading companies in this innovative field.

What is a Micro Data Center?

A micro data center (MDC) is a small, self-contained data center that is designed to be deployed in close proximity to the users or devices it serves. Despite its small size, a micro data center includes all the essential components of a traditional data center, such as servers, storage, networking equipment, power, and cooling systems.

Micro Data Center with Multiple Rack Units for Servers and IT Devices in an Enclosure

Micro data centers are portable and come in various sizes, ranging from a half-rack to a single rack, multiple racks, or even a small ISO shipping container (20 feet long by 8 feet wide). They can be deployed indoors or outdoors, depending on the specific requirements. While most micro data centers are floor-standing, smaller configurations can also be wall-mounted for added flexibility.

As illustrated below, micro data centers represent the smallest form of data center deployment. They typically support IT loads ranging from 1 kilowatt (kW) to 100 kW, distributed across 1 to 10 racks, and usually occupy less than 100 square feet of IT space.

Micro data centers are frequently utilized for edge computing applications, where low latency and localized data processing are critical. In this context, they are situated at the “far edge”, which refers to the outermost, decentralized locations of a network, nearest to the data source or end user devices, where data is initially generated and processed. These computing facilities cater to local markets by supporting various edge computing scenarios, including the Internet of Things (IoT) edge, on-premises edge, and tower edge.

Micro data centers are housed in a variety of locations, such as:

  • Commercial and Institutional Facilities: Retail stores, banks, small offices, hospitals, schools, airports, train stations, sports stadiums, and on the rooftops of commercial buildings
  • Industrial and Remote Locations: Manufacturing facilities, farms, and remote locations like oil and gas production sites
  • Telecommunications Infrastructure: At the base of cell towers, within Cloud-Radio Access Network (C-RAN) hubs, at cable headends, at the intersections of fiber optic routes, inside telecommunications rooms, and inside IT closets

Due to their versatility and varied applications, micro data centers are referred to by many different names. Some of these include micro edge data centers, micro-modular edge data centers, micro modular data centers (MMDC), mobile edge data centers, access edge data centers, containerized data centers, small data centers, mini data centers, micro DCs, MDC data centers, and data center-in-a-box.

Key Components of a Micro Data Center

The key components of a micro data center are housed within a secure, weatherproof enclosure that protects the equipment from environmental factors such as temperature fluctuations, humidity, and dust.

Key Components of a Micro Data Center with an Enclosure Displaying Network Gear on Wheels in a Room

Micro data centers typically integrate the following subsystems:

  • Rack Enclosure: Provides the physical framework to mount and organize IT and infrastructure equipment, ensuring efficient use of space and proper organization of components
  • IT Equipment: Core computing, data storage, and communications components are housed within the rack enclosure. These include servers, storage devices like SSDs (Solid State Drives) and HDDs (Hard Disk Drives), and networking equipment such as switches and routers
  • Cabling: To accommodate the compact size of micro data centers and their potential use in mobile or rugged environments, efficient and high-density fiber optic cabling is utilized
  • Uninterruptible Power Supply (UPS): Rack-mounted UPS systems provide both backup power during outages and continuous power conditioning to protect against voltage sags, spikes, and noise. The UPS system, including its associated batteries, serves the IT equipment, cooling systems, PDUs, security systems, and lighting with stable and reliable power
  • Power Distribution Units (PDUs): Rack PDUs, typically fed by the UPS, manage and distribute electricity to various IT devices within the micro data center
  • Cooling Systems: Micro data centers utilize various cooling solutions, depending on their design, which may include ventilation grilles, fans, air conditioners, or liquid cooling systems. For larger loads deployed using air-cooled systems, an external condenser may be installed on the roof or side of the facility housing the micro data center to supplement cooling capacity
  • Environmental Monitoring: Integrated, pre-wired sensors and Data Center Infrastructure Management (DCIM) software are used to monitor temperature, humidity, airflow, smoke, and flooding within the micro data center. These systems often include built-in digital displays, remote access monitoring capabilities, and the ability to make a simple connection to an organization’s utility power and network infrastructure
  • Physical Security: To protect the micro data center from unauthorized entry, physical security measures such as locks, biometric access control, and surveillance cameras are implemented

Advantages of Micro Data Centers

The main advantages of micro data centers are scalability and flexibility, reduced latency, cost-effectiveness, and rapid deployment.

1. Scalability and Flexibility

Micro data centers offer a modular approach to data center infrastructure, allowing organizations to scale their computing resources as needed. Like building blocks, these units can be easily deployed, removed, or relocated to adapt to changing business demands. This scalability enables companies to quickly expand or reduce their data processing capabilities without overprovisioning resources. By interconnecting multiple micro data center units, organizations can further expand their computing resources and capacity.

Another key advantage of micro data centers is their ability to be ruggedized. This involves reinforcing and waterproofing their physical structure with heavy-duty materials, such as welded steel frames, seals, and gasketing on the doors, thermal insulation, and robust casters. Ruggedization allows micro data centers to withstand harsh environmental conditions, including extreme temperatures, humidity, rain, flooding, dust, falling debris, and vibrations. This makes them particularly beneficial for industries that operate in remote or challenging environments, such as oil and gas, mining, and defense, where traditional data centers may not function effectively.

Ruggedized Micro Data Center with Strong Doors that are Closed and it Stands on Wheels in an Empty Room

Operating in a “lights-out” manner, micro data centers can function completely autonomously without the need for on-site human staff. This is crucial for systems that are being deployed in large numbers across wide geographies, as they can be managed remotely from a singular interface. The autonomous operation of micro data centers streamlines data center management processes and allows infrastructure to scale without the proportional need to increase staffing levels.

Example – Underwater Data Center

Microsoft’s Project Natick is a unique example of the flexibility of a micro data center (MDC), involving the deployment of small, self-contained data centers underwater. The project consisted of two deployments, each featuring a pre-fabricated, ruggedized, and completely waterproof containerized data center housing racks filled with servers. These micro data centers were designed to withstand the immense pressure found at ocean depths and utilize the naturally cold ocean water for cooling.

Microsoft Project Natick Phase 2 Underwater Data Center
Source: Microsoft.

The primary objective of Project Natick was to deploy these underwater data centers in close proximity to coastal urban centers. By doing so, Microsoft effectively implemented an edge computing solution aimed at reducing latency and improving the performance of cloud services for nearby populations.

2. Reduced Latency

Micro data centers can be placed in close proximity to the source of data generation or end users. This proximity significantly reduces the time it takes for data to travel (latency) between these small computing facilities and the data source or end user. While centralized data centers or cloud facilities typically have a minimum latency of 10 to 100 milliseconds (ms) when processing data, micro data centers can provide much lower latency, in the range of 1 to 10 ms or even sub-millisecond latency in some cases.

Micro Data Centers Situated in Close Proximity to Data Production of End Users
Source: Cellnex.

Reduced latency leads to faster round-trip times, resulting in quicker response times and improved performance for applications such as real-time analytics, Internet of Things (IoT) devices, and edge computing tasks. This enhanced response time improves user experience in various scenarios, including online gaming, video streaming, high-frequency trading, and autonomous vehicles.

3. Cost-Effectiveness

Micro data centers often require less upfront capital expenditure and ongoing operational expenses than traditional purpose-built data centers for several reasons:

  • Lower Initial Costs: The smaller size of micro data centers means both the setup costs and the real estate required are significantly less than the costs for traditional data centers
  • Flexible Expansion: Organizations can incrementally expand the physical space, infrastructure, and capacity of a micro data center (MDC) as needed in a “pay as you grow” model. This approach helps manage capital investments in a flexible manner
  • Reduced Power Consumption: Micro data centers typically consume less power than traditional data centers due to their smaller scale and more efficient design. Moreover, their modular approach allows power consumption to be closely aligned with actual demand, avoiding the underutilization of resources
  • Efficient Cooling Systems: Micro data centers utilize in-rack cooling systems, which are more targeted and effective in their approach to heat rejection. This eliminates the need for larger, more energy-intensive room-level cooling solutions
  • Simplified Maintenance: Standardized components in micro data centers can be replaced by non-technical staff, reducing both the expertise required for maintenance and its associated costs

In addition to these facility-level cost savings, micro data centers can also reduce bandwidth costs by keeping data traffic localized. By processing data closer to the source, less data needs to be transmitted over long distances to and from centralized data centers or the cloud. This helps save on third-party network resource usage and reduces network congestion, which is particularly important for bandwidth-intensive applications.

Another benefit of micro data centers is the potential for tax savings. In many jurisdictions, these units are categorized as equipment (personal property) instead of building improvements (real property) for tax purposes. This classification allows companies to depreciate the cost of the micro data center more rapidly compared to traditional data center buildings. Faster depreciation results in a quicker recovery of the investment through lower taxable income in the early years of ownership.

4. Rapid Deployment

Micro data centers are pre-configured and pre-tested units that can be quickly deployed and commissioned, often without requiring specialist technical knowledge. This allows for easy installation in various environments, significantly reducing deployment time compared to traditional data centers. While constructing a traditional data center can take several months, micro data centers can be manufactured and shipped within weeks and installed within days.

The ease of deployment for micro data centers is particularly beneficial for businesses expanding their operations into remote or temporary locations that lack the infrastructure for traditional data processing facilities. Two examples of such locations are:

  1. Remote: Oil and gas exploration sites can utilize micro data centers to process and analyze data obtained from sensors and monitors for seismic surveys, drilling operations, and pipeline monitoring
  2. Temporary: Shopping malls can install a micro data center in an IT closet to support additional kiosks during the holiday shopping season, and then remove them when the holiday season is over

The compact, self-contained design of micro data centers, with integrated power, cooling, and networking equipment, makes them highly portable. This all-in-one enclosure enables them to be easily and rapidly disconnected, transported, and re-installed in new locations without the need for the extensive infrastructure build-out required for a traditional data center.

Example – Formula 1 Racing

The 2024 Formula 1 calendar features Grand Prix events in 21 countries across five continents, underscoring the global scale of the sport. This widespread nature requires rapidly deployable IT infrastructure, making micro data centers an ideal solution. Their portability allows for quick setup and dismantling of the necessary computing resources at each race location, ensuring efficient operations throughout the season.

Formula 1 distinguishes itself from other sports through its heavy reliance on real-time data and computational analysis. Each race car is equipped with over 300 sensors, generating more than 1.1 million data points per second. This data provides vital information for optimizing race strategy and car performance, making a data-driven approach crucial for success in the sport.

Holographic Formula One Car Speeds On Dark Track Grid with Light Trails Behind

At the racetrack, micro data centers play critical roles both before and during the race. Before the race, they enable teams to process real-time weather data and forecasts, assisting in making critical decisions regarding tire choices and car setups. During the race, these micro data centers facilitate the analysis of real-time data on key factors such as tire degradation rates, fuel loads, track conditions, and the positions of competitors. This information is crucial for calculating the optimal timing for pit stops and adapting race strategies on the fly.

Use Cases of Micro Data Centers

Micro data centers are crucial for enabling edge computing capabilities. They serve as nodes for various applications, such as the Internet of Things (IoT), real-time analytics, and Content Delivery Networks (CDNs). Additionally, micro data centers support the infrastructure requirements of 5G telecommunications networks and provide IT services for remote and branch offices.

1. Edge Computing Applications

Internet of Things (IoT)

Internet of Things (IoT) networks, which are comprised of sensors and connected devices, generate vast amounts of data that often requires real-time processing and analysis. These devices are located in diverse settings, including homes, urban environments, industrial sites, manufacturing facilities, agricultural fields, and healthcare facilities.

Internet of Things IoT Connections shown in a City at Sunset with Data and Digital Communication Nodes

Moreover, IoT connections are growing rapidly – Ericsson’s Mobility Report projects a significant expansion of global IoT connections, from 15.7 billion at the end of 2023 to 38.9 billion by the end of 2029, representing a 16% compound annual growth rate (CAGR).

Micro data centers play a critical role in IoT ecosystems by providing localized computing and storage resources closer to IoT devices. By processing data near its point of generation, micro data centers can support:

  • Time-Sensitive Applications: Immediate data processing near IoT devices facilitates faster response times and enables autonomous decision-making. This is essential for time-sensitive applications like autonomous vehicles, emergency response systems, industrial automation, and remote surgery
  • Data Filtering and Preprocessing: Micro data centers can filter and locally preprocess large volumes of raw IoT data before sending relevant information to centralized data centers or the cloud. This reduces unnecessary network traffic, saving on bandwidth costs and lowering data storage costs

Manufacturing and industrial environments are increasingly adopting Industrial Internet of Things (IIoT) devices and sensors for monitoring and optimizing production processes. Being ruggedized and close to data sources like machinery and production lines, micro data centers process the vast amounts of data generated by these IIoT devices in real-time. They facilitate the analysis and use of data for operations such as quality control, predictive maintenance, supply chain optimization, and automation.

Real-Time Analytics

Real-time analytics requires high-speed data processing to extract insights from data as it is generated by sensors, machines, or user devices. This is crucial in many scenarios, such as financial trading, patient monitoring, and supply chain management.

Man Contemplating Real-Time Analytics Dressed In Blazer with Data On Multiple Screens in Front of Him

Micro data centers enable businesses to achieve the following:

  1. Immediate Insights: By bringing processing power closer to the data source, micro data centers allow businesses to perform real-time analytics on live data streams. This enables the identification of patterns, anomalies, and trends in real-time, which is vital for applications that require informed, split-second decisions, such as fraud detection, predictive maintenance, or real-time recommendations
  2. Low-Latency Decision Making: Micro data centers minimize the distance data needs to travel from its source to its destination, thereby reducing the time delay (latency) in the responsiveness of analytics and the decision-making process. Centralized data centers or cloud-based solutions typically have a minimum latency of 10 to 100 milliseconds (ms) for processing, which is unacceptable for many use cases. In contrast, micro data centers can reduce this latency to be in the range of 1 to 10 ms, or even sub-millisecond latency in some cases. This allows businesses to make faster, more informed decisions, ultimately improving operational efficiency and customer experiences

Content Delivery Networks (CDNs)

Content Delivery Networks (CDNs), operated by companies such as Akamai, Cloudflare, Fastly, and Edgio, are designed to deliver content to users with high availability and performance. Micro data centers play a crucial role in CDN architectures by being deployed in close proximity to end users. This strategic placement significantly enhances content delivery in the following ways:

  1. Reduced Load Times: Micro data centers function as edge caching points within CDNs, storing popular content such as websites, images, videos, audio files, and software updates closer to end users. By decreasing the distance data must travel, these micro data centers facilitate much faster loading times compared to retrieving content from distant, central servers. This proximity translates into a smoother user experience, particularly when accessing high-bandwidth content like streaming video
  2. Optimizing Network Traffic: By placing micro data centers at the edge of the network, CDNs can route and backhaul traffic more efficiently. This optimization reduces strain on the core network and mitigates bottlenecks, providing smoother content delivery even during peak times. Frequently requested content is cached and served directly from this edge data center, minimizing traffic back to the origin servers. This approach not only improves the overall performance of the CDN but also enhances cost efficiency by reducing the load on central infrastructure

2. Telecommunications – 5G Networks

Micro data centers are being placed at network aggregation points to enable the full potential of 5G networks by supporting low-latency, high-bandwidth applications through localized data processing and offering flexible, scalable, and cost-efficient hosting for virtual network functions (VNFs).

5G Cell Tower Radiates Frequencies Amongst Trees Under Starry Night Sky in Rural Area

Speed and Latency

5G networks aim for significant enhancements compared to 4G, targeting improvements in average download speeds by 10 to 100 times, reducing average round-trip latency by 5 to 10 times, and increasing the maximum number of simultaneous connections per cell by up to 300 times.

To achieve these enhancements, 5G networks are strategically placing micro data centers closer to the network edge, such as at the base of cell towers and in nearby Centralized Radio Access Network (C-RAN) hubs.

Connections to Regional Data Center via Fiber Links to Micro Compute Facility and Enterprise Building
Source: American Tower.

By bringing computing resources closer to the source of data generation, this proximity enables low-latency and high-bandwidth applications, such as high-resolution video streaming. Processing data near its origin reduces the distance it needs to travel, thereby minimizing latency and improving overall performance.

Network Functions Virtualization (NFV)

Network Functions Virtualization (NFV) is a technology that enables the transition of network functions from traditional, proprietary hardware appliances to virtualized software applications running on general-purpose servers. This shift is particularly advantageous for 5G networks, as it allows for greater flexibility, scalability, and cost-efficiency compared to hardware-based network infrastructure.

Micro data centers play a crucial role in hosting these virtual network functions (VNFs), such as routers, firewalls, and load balancers. These small data centers can be deployed quickly in various distributed locations, positioned within a few milliseconds of the Remote Radio Heads (RRHs) located on cell towers. The servers housed in these micro data centers support virtual Baseband Unit (BBU) functionality, which is essential for the operation of 5G networks as it enables the centralization and virtualization of radio access network (RAN) functions.

By leveraging NFV and deploying VNFs in micro data centers, telecommunications operators can optimize their 5G network infrastructure and reduce their reliance on expensive proprietary hardware. Furthermore, these micro data centers serve a dual purpose – not only do they host the VNFs, but they also provide sufficient server capacity to run the cloud services that deliver applications at the edge over the 5G network.

READ MORE: Network Functions Virtualization (NFV) Explained

3. Remote and Branch Offices

Micro data centers provide a solution for delivering critical IT infrastructure, such as servers, storage, and networking, to remote or branch offices without requiring extensive on-site IT space and personnel. By hosting applications and data locally, micro data centers improve performance and reduce latency compared to centralized data centers or cloud-based solutions. These standardized deployments provide business continuity in the event of network disruptions or failures at the main data center, allowing remote offices to continue operating seamlessly.

Micro Data Center Companies

Micro data center companies can be categorized into two main groups: operators and manufacturers. Micro data center operators manage and maintain the infrastructure and services of these computing facilities, ensuring their smooth operation. On the other hand, micro data center manufacturers focus on designing and building the physical infrastructure and components of these self-contained units.

Micro Data Center Operators

Micro data center (MDC) operators often deploy these facilities using a colocation model, where multiple customers can rent space, power, cooling, and connectivity for their IT equipment within the same facility. This shared infrastructure is frequently located at the base of cell towers and in nearby C-RAN (Cloud-Radio Access Network) hubs. Some notable micro data center providers include:

  • American Tower: Owns multiple micro data centers deployed at the base of its cell towers across the United States. Each facility comprises approximately 360 square feet and houses over 20 customer cabinets per location. CoreSite, American Tower’s data center operations subsidiary, manages these facilities
  • SBA Communications: Through its SBA Edge offering, the company has deployed micro data centers at the base of its cell towers. SBA has stated that it has over 8,000 tower sites in the United States capable of housing a micro data center comprising between 6 and 20 racks, with secure and sizable ground space, adequate power availability, and fiber connectivity
  • Ubiquity Edge: Formerly known as EdgePresence, the company provides modular and purpose-built micro data centers designed for carriers and hyperscalers in multiple markets across the United States. Their facilities prioritize efficiency in critical power, monitoring, physical security, and cooling
  • Vapor IO: Specializes in edge computing, offering micro data center and networking services designed to support applications requiring low latency and high bandwidth. Crown Castle, an investor in Vapor IO and a key partner, supplies real estate and fiber connectivity to support Vapor IO’s Kinetic Grid platform

Ultimately, system integrators, IT service providers, and end users are responsible for deploying the IT equipment and applications that operate inside these micro data centers.

Micro Data Center Manufacturers

Micro data center (MDC) manufacturers are global companies that provide a range of products and services for power management and cooling systems across various critical industries, with computing facilities being a key area of focus. Some of the leading manufacturers in this field include:

  • Eaton: A global power management company that manufactures micro data centers and offers them under the Tripp Lite brand. Eaton provides integrated solutions that combine power distribution, power quality, and backup power into compact and efficient systems
  • Schneider Electric (including its subsidiary APC): As a global specialist in energy management and automation, Schneider Electric offers pre-configured, fully integrated, and self-contained micro data centers branded as EcoStruxure. These solutions include power, cooling, security, and management components
  • Vertiv: Designs, builds, and services critical infrastructure for data centers. They offer a wide range of micro data center products, such as self-contained, rack-level systems with integrated power, cooling, and management features
  • Zella DC: Has supplied micro data centers as part of an edge computing solution for Internet Initiative Japan (IIJ), targeting telecom and 5G customers. These facilities have been deployed in Japan and Iceland

Cost of Micro Data Centers

The cost of a micro data center (MDC) will vary between $10,000 to over $1 million based on its form factor, specifications, redundancy levels, integrated technology, and the intended deployment environment.

Mini Micro Data Center with Doors Open and Lights On Displaying Servers IT Equipment Data Center

The price range for different MDC configurations is as follows:

  • Single Rack: Costs typically range from $10,000 to $75,000
  • Multiple Racks: Increasing the facility size to accommodate 5 to 10 racks raises the cost to $100,000 to $500,000, depending on the technology and capacity involved
  • Larger Configurations (Shipping Container): Prices can reach $1 million or more, reflecting the increased capacity, power, cooling, and networking features
Mary Zhang covers Data Centers for Dgtl Infra, including Equinix (NASDAQ: EQIX), Digital Realty (NYSE: DLR), CyrusOne, CoreSite Realty, QTS Realty, Switch Inc, Iron Mountain (NYSE: IRM), Cyxtera (NASDAQ: CYXT), and many more. Within Data Centers, Mary focuses on the sub-sectors of hyperscale, enterprise / colocation, cloud service providers, and edge computing. Mary has over 5 years of experience in research and writing for Data Centers.

LEAVE A REPLY

Please enter your comment!
Please enter your name here