This article is based on an interview originally published by Digital Journal with Arun Shenoy, VP at ServerFarm, discussing how cloud computing is transforming data center operations and strategy.
The Evolving Role of the Data Center
Cloud computing has fundamentally changed how businesses deploy applications, manage users, and deliver resources. Rather than operating isolated on-premise infrastructure, organizations can now provision virtual data centers in the cloud — gaining greater compute power, improved data security, and more flexible access to tools and services.
Modern data centers are complex systems composed of routers, switches, security appliances, storage systems, servers, and application delivery controllers. The ability to interconnect these elements — whether on-premise, in a colocation facility, or across public cloud providers — is enabling organizations of all sizes to become more agile and cost-competitive.
How Fast Is Data Center Traffic Growing?
According to Arun Shenoy, VP at ServerFarm, global data center traffic is growing substantially as cities, workplaces, and individuals depend more heavily on connected technologies for everyday activities. The drivers are varied: social media users uploading content, employees sharing files across distributed teams, and the rapid proliferation of Internet of Things (IoT) devices are all compounding demand for storage and bandwidth.
Shenoy cited the Global Interconnection Index, which projected that data center interconnection bandwidth would grow at approximately 45 percent per year between 2016 and 2020. While that specific forecast period has passed, the underlying trend — accelerating demand for interconnected data center capacity — continues to shape infrastructure investment decisions today.
Four Key Drivers of Data Center Expansion
Shenoy identified four primary forces driving organizations to expand their data center footprint:
1. Cloud Migration of Applications and Workloads
The cloud is the single most significant driver behind data center expansion. Moving workloads to cloud infrastructure allows businesses to scale faster, iterate more quickly, and run a broader range of applications with lower capital overhead. As more workloads shift from on-premise servers to cloud platforms, the demand for robust, interconnected data center infrastructure grows in parallel.
2. The Shift to Mobile Infrastructure
Mobile devices — smartphones, laptops, wearables, tablets — each carry significant compute and connectivity requirements. The rollout of 5G networks amplifies this further, enabling faster data transfer rates and supporting a higher density of connected devices. This increased mobile traffic places greater demands on backbone infrastructure, making larger and better-optimized data centers necessary to maintain acceptable performance.
3. The Rise of the Internet of Things (IoT)
Traditional internet usage focused on connecting applications to people — social media, collaboration tools, document sharing. IoT extends this model: devices talk to devices, people interact with machines, and machines generate continuous telemetry streams that require storage, processing, and analysis. Shenoy noted that while earlier estimates counted around 8.5 billion connected mobile devices, adding IoT endpoints could push that number to 50 billion or more. That device explosion creates enormous downstream pressure on data center capacity and efficiency.
4. Data Sovereignty and Regulatory Compliance
Regulatory frameworks like the EU’s General Data Protection Regulation (GDPR) require organizations to be deliberate about where data is stored and how it is processed. As businesses operate across borders, they must often maintain data within specific geographic jurisdictions. This drives investment in regionally distributed data center infrastructure and complicates simple “move everything to the cloud” strategies.
The Business Case for Cloud Computing
The advantages of cloud computing for most organizations come down to three attributes: speed, flexibility, and cost efficiency.
- Speed and agility: Cloud platforms allow teams to provision infrastructure in minutes rather than weeks, accelerating development cycles and enabling rapid iteration.
- Location independence: Because applications and data reside in cloud data centers rather than local machines, employees can access business systems from any location with an internet connection — a capability that has proven critical for distributed and remote teams.
- Pay-as-you-go economics: Rather than purchasing and maintaining expensive on-premise hardware, organizations pay only for the compute and storage they consume. Host vendors manage hardware maintenance, power, and cooling, shifting significant operational overhead off the customer’s balance sheet.
Will Cloud Computing Replace Traditional Data Centers?
The short answer, according to Shenoy, is no — at least not entirely. The growth of cloud services has driven the rise of hyperscale data centers operated by major cloud providers, which are optimized for enormous scale and automation. Consumer-facing applications like video streaming, gaming, and social networking have accelerated this trend, as have enterprise workloads running ERP, CRM, and analytics platforms.
However, many organizations are adopting a hybrid approach rather than fully migrating to the cloud. IoT use cases — smart city infrastructure, connected vehicles, digital health devices — often require edge processing and on-premise components alongside cloud integration. The result is that traditional and cloud data centers are increasingly complementary rather than competitive.
On-Premise vs. Cloud: It Depends on the Organization
Not every workload is a good candidate for cloud migration. Shenoy highlighted several industries — healthcare, pharmaceuticals, power utilities, and government — where regulatory requirements, performance constraints, or data sensitivity make full cloud migration impractical or prohibited. Organizations in these sectors typically maintain a hybrid posture, keeping sensitive or regulated workloads on-premise while leveraging cloud for appropriate use cases.
The decision framework is rarely purely technical. Country of operation, applicable data residency laws, latency requirements, and vendor risk tolerance all factor into where a given workload should run.
Security Considerations: Cloud and On-Premise
Both deployment models carry distinct security challenges that DevOps and infrastructure teams need to address deliberately.
Cloud Security
When moving to the cloud, organizations are effectively placing their data and assets in infrastructure they do not physically control. While this enables cost savings and operational flexibility, it introduces a shared responsibility model: the cloud provider secures the underlying infrastructure, but the customer remains responsible for access controls, data classification, and application-layer security. Shenoy emphasized the importance of thoroughly vetting cloud vendors and understanding the contractual and technical boundaries of responsibility before committing sensitive workloads.
On-Premise Security
On-premise data centers present their own risks, particularly around legacy operational technology (OT). Cooling systems, power management units, and building management systems are often aging infrastructure that has only recently been connected to the internet. This connectivity, while enabling remote monitoring and management, also introduces new attack surfaces. Shenoy noted that power failures are a common trigger for opportunistic attacks, and that securing these OT systems — while simultaneously gaining visibility into them — requires specialized expertise.
The Future: Intelligent, Self-Managing Infrastructure
The trajectory of cloud data center development points toward increasing automation and intelligence. Shenoy observed that cloud environments are already more automated than traditional IT environments, with fewer human operators managing larger pools of infrastructure. The next phase involves applying AI and machine learning to create self-managing infrastructure — systems capable of detecting anomalies, optimizing resource allocation, and responding to failures with minimal human intervention.
For DevOps practitioners, this has concrete implications: the operational model for infrastructure is shifting from reactive administration to proactive policy definition. Engineers define the desired state and guardrails; increasingly intelligent systems handle moment-to-moment execution.
Key Takeaways
- Cloud computing is not eliminating data centers — it is transforming their role and architecture.
- Four forces are driving data center expansion: cloud migration, mobile growth, IoT proliferation, and data regulation.
- Most enterprise organizations are adopting hybrid approaches rather than full cloud migration.
- Both cloud and on-premise deployments carry distinct security responsibilities that must be actively managed.
- The long-term direction of data center infrastructure is toward greater automation, AI-driven management, and self-healing systems.