In IT, there’s seems to be a tendency to toss around the latest buzzwords without actually unpacking their nuances. The terms “VDI” and “desktop virtualization” are often used interchangeably, while the distinction between various kinds of virtualization (desktop, server, storage, network, and more) is sometimes ignored.
Comparing desktop and server virtualization is like comparing apples and oranges. Yes, the epic war between two fruits that has raged on for generations with no end in sight. And when you consider all the food groups out there, you know they actually have a few things in common (just don’t tell the fruits that).
When we zoom in, it’s easy—and important—to highlight some key differences. Server virtualization means (you guessed it!) virtualizing physical servers in the data center, while desktop virtualization means virtualizing the desktops of end users in the data center. In both cases, we’re adding an abstract layer–often referred to as the virtualization layer of hypervisor–to our physical server.
With server virtualization, we are slicing the server into containers (abstractly, of course). This can happen one of two ways: that layer can either go directly on the hardware, which is referred to as bare metal, or it could be hosted via the operating system environment. When it’s the latter, the operating system is a common denominator, also being sliced into containers. Afterwards, different applications can be added and executed in each isolated virtual environment.
In both cases, the central server is running a number of images, usually with the goals of better utilizing workloads, better dealing with variance in workloads, and/or testing a new environment. Imagine a bunch of different workloads coming together like a puzzle. When you think of server virtualization, you should think of consolidation.
With desktop virtualization, on the other hand, you should think of separation. Every end user has its own desktop operating system and applications. But instead of being stored on the hard drive of that user’s personal device(s), each desktop runs in a virtual machine on a server.
Whether it’s through a laptop, desktop, tablet, smartphone or other device, users can then easily and quickly access their virtual machine (which, once again, is stored in the data center) from anywhere with a simple login.
Centralization is a key benefit of desktop virtualization. That’s because, with desktop virtualization, there’s one “golden image” from which additional virtual desktops are cloned from. Naturally, this makes provisioning new desktops significantly faster and makes it less concerning when an employee spills apple juice (or orange juice—we’re not picking sides) all over their laptop.
These desktops can be personalized with different applications, but the central golden image is important. When software updates are needed, for example, you can just change the one “golden image” instead of having to change every machine individually. It’s that simple.
Please note: Using a thin client is technically not desktop virtualization, because there is no hypervisor. Yet because the process is similar (although graphics quality and application runtimes are significantly lower), many refer to it as desktop virtualization anyway.
If you’re still confused about the distinction between the two, we like this summary from ZDNet: “Desktop virtualization means that you run a virtual machine on your desktop computer. Virtual desktop infrastructure (VDI) is a data center technology that supplies hosted desktop images to remote users.”
Add it up, and desktop virtualization gives IT far more control and security over the environment without sacrificing user freedom. And it’s the foundation for trends that have been “cloud-washed” lately, like BYOD (bring your own device), secure remote access, and others.