The cocky 20-something me used to think I understood compute.
Then, during my first program right after my Master's, I managed the hardware deployment of hundreds of physical servers. My team built them into clusters to migrate thousands of VMs from an old legacy platform to newer infrastructure.
And throughout the entire project, it hit me hard: I didn't understand compute as well as I thought.
I could tell you what a VM was. I could tell you why we used them. But if you asked me to explain the layers—hardware, hypervisor, VM, application - and how everything fits together, I’d struggle pretty badly.
These are the resources that helped me connect the dots back then.
Resource 1: Virtual Machines vs Containers
What it is: A short and sweet video that tells the story of the evolution of compute. It starts with the world before virtual machines, where physical servers meant wasted money and wasted capacity. Then it explains why people moved to virtual machines to optimize compute resources. It then covers the drawbacks of virtual machines and explains how people pivoted to a new architecture called containers.
Why it helped me: This was one of the first resources that helped me connect the dots on what I was actually working on during that cluster migration program. I love how the video explains the architecture layer by layer—hardware, hypervisor, VM, application—and suddenly all the pieces I'd been working with clicked into place. I finally understood why we were deploying hardware, building clusters and ESXi hosts, and how that mattered to the VM migration that we were doing.
Who this is for: Anyone who's heard "virtual machine" and "container" a hundred times across random meetings but isn't sure how they're different, or why one exists when the other does too.
Skip this if: You already live and breathe VMs and containers on a daily basis. This video is beginner stuff, perfect for a nerdy project manager like me.
Pro tip: Pay attention to the drawbacks of physical servers and virtual machines. Once you understand those, picking the right compute options for your use case starts to make a lot more sense.
Resource 2: Hyperconverged Infrastructure (HCI) Explained in 5 Minutes
What it is:
If you have been hearing about Hyperconverged Infrastructure (HCI) but have absolutely no clue what it is, this video is a good start. Instead of deploying and managing three layers of infrastructure hardware (compute, networking, and storage) separately, HCI bundles them into a single system with the “integration” happening at the software layer to simplify deployment and maintenance.
HCI isn’t a single product from a single vendor. Nutanix, VMware, Dell, HPE, and Microsoft each offer their own range of HCI solutions.
Why it helped me:
In my line of work, I have seen both the traditional three-layer IT infrastructure and Hyperconverged Infrastructure in action. This video nicely filled a gap in my understanding of the differences between the two types of infrastructure.
I walked away with a gentle reminder that HCI doesn't eliminate hardware. It eliminates the complexity of managing separate hardware layers.
The storage that used to live in a dedicated SAN array now lives on local disks inside each server node. A software layer (like Nutanix's AOS or VMware vSAN) pools those disks together and presents them as shared storage.
Who this is for: Anyone who's heard "HCI" or "hyperconverged" in meetings but wasn't sure what it actually meant, or how it even relates to compute.
Skip this if: You're after a comprehensive guide for deployment. This explainer video focuses on the theory. Deploying HCI at scale is a totally different story—and it would take serious work to get your hands dirty. Not just a 5-minute video.
Pro tip: The contrast between the "Traditional IT Architecture vs. HCI" makes it clear why HCI became popular and where it fits in the evolution from physical servers to virtual machines to hyperconverged infrastructure.
Resource 3: Bare Metal vs Virtual Machine vs Container
Source: NetActuate - Bare Metal vs Virtual Machine vs Container: The Complete Infrastructure Comparison
What it is: If you’re into reading, this is a balanced, practical comparison of the three compute options: bare metal, virtual machines, and containers. Instead of dismissing bare metal as "old furniture" or treating cloud as the inevitable answer, the article frames the decision around one question: which compute layer does this workload belong on? The use-case matrix covering eight production workload types is gold as it nicely explains where each tier is the optimal choice.
Why it helped me:
I've seen all three types available and being used at a financial institution. But I never quite understood how the decisions were made on which one goes where.
This piece offers a clear, honest explanation with a balanced viewpoint: bare metal wins on raw performance and security isolation. VMs win on resource efficiency and workload flexibility. Containers win on deployment speed and portability. Essentially, no one-size-fits-all solution exists, so match the workload to its dominant constraint.
Bare metal for database primaries, media encoding, and AI training. VMs for management tooling, DNS, and compliance-sensitive multi-tenant environments. Containers for APIs, microservices, and CI/CD pipelines. It makes a lot more sense to me now.
Who this is for: Anyone who's seen bare metals, virtual machines and containers in production but never had a clear framework for why one was chosen over another.
Skip this if: You're looking for a beginner's introduction to what these three things are. This can get a bit too technical in the explanation and assumes you already know the basics and want to understand the trade-offs between each option.
Pro Tip: A common misconception is that containers are replacing VMs or bare metal. They’re not. Containers don’t run from thin air. They still need VMs or bare metal to run on.
Resource 4: Data Centre Consolidation Calculator
Source: Robert Frances Group - Data Centre Consolidation Calculator
What it is: An interactive calculator that models the real economics of infrastructure decisions. You can punch in the details about your current fleet and a target modern platform, and it estimates the consolidation ratio, recovered power, space, licensing, and net TCO impact over your analysis term. Bonus: it also includes a breakdown of where annual savings come from—power and cooling, software licensing, maintenance and support—and shows cumulative net position over three or five years, depending on your preferred analysis term.
Why it helped me: Infrastructure is usually viewed as a cost centre, and infrastructure decisions are rarely a technical discussion. They're investment decisions with ROI to discuss, numbers to defend, and budgets to approve.
Seeing what’s required to model the net Total Cost of Ownership (TCO) offers a unique lens into what kind of tough questions to ask, how much homework is required to translate the vendors’ quotes, the data centre’s cost estimates, the contractual terms and the intended benefits into dollars and cents that can persuade the board to put more money on the line.
If you can’t make sense of the numbers yourself, with facts and assumptions you can defend, there’s very little chance your charisma alone will get the board to sign off on the investment.
Who this is for: Anyone who needs to justify a compute decision to finance or who wants to see for themselves why the cheapest option isn't always the right one.
Skip this if: You're looking for a conceptual explanation of bare metal, VMs, and containers. This is a financial modelling tool, not an explainer.
Pro Tip: The default licensing rate is modelled per core, but most enterprise contracts are structured differently. So question and play hard with the assumptions. Adjusting that rate is where the real insight lives.
Where to Start
If you're completely new to compute, start with Resource 1. It gives you the most beginner-friendly narrative on how things start and why physical servers gave way to virtual machines and containers.
If you want to see how compute is deployed, watch Resource 2. This one will give you insight into how HCI eliminates the complexity of managing separate hardware layers, simplifying integration and maintenance at scale for real-world organizations.
And when you’re ready to dip your toes a bit deeper into the water, Resource 3 lays out the trade-offs clearly and explains how to decide which workloads to align with which compute options.
And for those looking to reimagine and overhaul your infrastructure fleet, Resource 4 helps you move beyond technical considerations to the arguably tougher financial discussions needed to justify your proposed investment to the board.
Your Turn
What was your "aha" moment with virtual machines or containers? Or are you still in the "what even is this" phase?
I'd love to hear what questions are still puzzling you.
If you found this useful, subscribe to SkyeWorks. I write about technology the way it should be—simple, grounded, and human. No jargon, no ego, just guidance that actually helps.
See you in the next post.
