Chimezie Onwuka - Episode 166 - The Route to Networking

From Physics To The Data Centre Floor: Chimezie Onwuka On Building, Operating and Scaling Critical Infrastructure

AI, cloud computing and digital services continue to reshape the global economy, but beneath every application, transaction and AI model sits an often-unseen foundation: the data centre. In this episode of The Route to Networking podcast, host Tom Jones speaks with Chimezie Onwuka, an Uptime-certified data centre specialist with nearly a decade of hands-on experience designing, building and operating mission-critical facilities across multiple industries.

From physics labs to live production environments, Chimezie brings a grounded, real-world perspective on how data centres work, why they matter more than ever, and what engineers need to thrive as AI, power constraints and scale reshape the industry.

The Data Centre as the Foundation of the Digital Economy

For Chimezie, the data centre is far more than racks, cooling units and power feeds. It is the bedrock on which modern enterprise and the digital economy are built.

“I like to see the data centre as the base of a pyramid on which enterprise is built. It’s the room where the mission statement of a company actually lives.”

He explains that without resilient, well-designed facilities, digital commerce, cloud services and global connectivity would simply not function. Over the past decade, as demand has surged and availability expectations have tightened, the margin for error inside a data centre has shrunk dramatically.

“Every decision, every action in a data centre has business effects. It’s a production environment – you don’t get to be casual.”

From Physics to the Data Centre Floor

Chimezie’s route into data centres was anything but conventional. Originally studying physics in West Africa, he quickly realised that while theoretical concepts such as quantum mechanics held little appeal, applied electronics and hands-on problem‑solving did.

“What I really enjoyed was being in the lab – figuring out how to connect things, how to calculate capacitance, how to work on boards. That’s when I realised this was pointing somewhere.”

An elective module in computer programming further cemented that shift, introducing him to Linux and systems thinking. While his degree helped him understand how he learned, he knew it would not be enough to break into live infrastructure environments.

That awareness led him to an intensive IT infrastructure bootcamp, focused on Linux administration and server environments. The decision proved pivotal, directly leading to his first role working inside a live data centre.

Breaking Into Data Centres Without a ‘Perfect’ Background

One of the strongest themes in the conversation is accessibility. Chimezie is clear that data centres are not reserved for one academic path.

“In my bootcamp class, we had people who studied zoology, law – all sorts. At that point, we’d all decided this is what we wanted.”

He encourages aspiring engineers not to limit themselves based on their degree, or lack of one. Instead, he emphasises adaptability, commitment and self-awareness.

“Do not box yourself in by what you studied. If you’re agile enough to reshape your skills, there is always an opportunity.”

Why You Must Understand What Sits on Top of the Data Centre

Chimezie stresses that effective data centre engineers must look beyond racks, power and cooling. The facility is only layer one.

“It’s not just racking and stacking. You should understand what’s built on top of the data centre – operating systems, applications, networks.”

That broader understanding allows engineers to troubleshoot intelligently, identify whether issues originate at the physical layer or higher up the stack, and communicate more effectively with other technical teams.

“If someone is talking about deploying applications behind load balancers across multiple data centres, you should already understand what that means.”

Learning the Industry From the Inside

Across his career, Chimezie has worked in managed services, telecoms, colocation and fintech – each environment shaping his approach.

In managed services, infrastructure could be powered down outside business hours. In telecoms, structure, cabling discipline and redundancy became critical. In fintech, downtime was simply not an option.

“Data centres have to be built with a clear understanding of the business they are supporting. If they’re not aligned, you either overspec or take unacceptable risk.”

Data Centre Tiers, Certification and Common Mistakes

Chimezie provides a clear, practical explanation of data centre tiers and why certification matters.

“A design can be Tier III on paper, but the constructed facility might not meet that standard. Certification has to start at design, not after the build.”

He outlines the progression from Tier I through to Tier IV, explaining concepts such as redundancy, concurrent maintainability and fault tolerance, and warns against treating certification as a marketing label rather than a technical discipline.

“If a colocation provider tells you they’re Tier III, ask for proof. Not words – evidence.”

AI, Power Density and the Cooling Challenge

Few forces are reshaping data centres faster than AI. Throughout the conversation, Chimezie repeatedly returns to the sheer scale of change driven by GPU-heavy workloads and machine learning models.

“We’re not talking CPU anymore. It’s GPU everywhere, and the power requirements are almost ten times what we used to know.”

Unlike traditional enterprise workloads, AI environments push racks far beyond historical density assumptions. This has direct consequences for power delivery, cooling strategy and overall facility design.

“It’s not just about cooling the server anymore. Heat dissipation itself becomes a design problem.”

Chimezie explains how unmanaged airflow and poor thermal planning can lead to demand fighting, where hot and cold air mix within the white space. This forces cooling systems to operate continuously at high load, driving up energy use and operational cost.

“When that happens, your cooling units are running at peak all the time. That’s bad for efficiency and bad for business.”

As a result, liquid cooling is rapidly moving from niche to necessity. For high-density AI racks, it offers a more efficient way to remove heat, reduce strain on CRAH units and keep power utilisation effectiveness within acceptable limits.

“Liquid cooling is quickly becoming the default. Air alone just isn’t enough anymore.”

Power, Sustainability and the Reality of Scale

While data centres can be built in months, power infrastructure often takes years – a growing constraint on industry expansion.

“You can build a data centre in a year. It can take five years to build the power plant.”

As facilities consume energy at city‑level scale, sustainability is no longer optional. Operators, regulators and governments are under increasing pressure to balance digital growth with environmental responsibility.

“We still need to live in our homes. There has to be a balance.”

Design and Operations: Two Sides of the Same Coin

One of the most practical insights Chimezie shares is the risk of treating design and operations as separate worlds.

“One mistake people make is leaving operations out of the design conversation. They’re the ones who live with the decisions.”

While design often focuses on specifications and certifications, operations teams are responsible for keeping environments running day after day. Decisions that look fine on paper can quickly become painful in practice if maintainability, access and real-world workflows are overlooked.

The strongest data centres, he explains, are those where operations, security, networking and compliance teams are involved early, before anything is signed off.

“If you don’t think about how something will be maintained, you haven’t finished designing it.”

Bringing design and operations closer together isn’t just an organisational consideration; it directly shapes the kind of engineers the industry now needs.

Advice for the Next Generation of Engineers

Chimezie’s advice to aspiring data centre engineers is grounded and direct:

“Try to be a balanced engineer. Understand operations and design. Know what’s built on top of the data centre.”

He also stresses the importance of curiosity and adaptability in an industry that evolves at speed.

“Things move fast – sometimes at the speed of light. If your job starts to feel routine, something is wrong.”

A Quick‑Fire Round Worth Listening To

The episode closes with our quick-fire round where Chimezie reflects on career-defining moments, emerging technologies and the future of AI-driven infrastructure.

Without giving anything away, he shares his thoughts on:

• The experience that most shaped his confidence as an engineer
• The emerging technologies everyone should understand in the next two years
• Why AI and machine learning are applying unprecedented pressure to infrastructure
• The mindset required to thrive in a fast-moving, high-stakes industry

🎧 Listen to the full episode of The Route to Networking podcast to hear Chimezie’s insights in his own words and explore what it really takes to build and operate the infrastructure powering today’s digital world.

Connect with Chimezie on LinkedIn here.