There's nothing new under the sun -- except maybe hyper-converged computing software.
When technology gains traction, gets lip service or becomes trendy, it's often passed off as something new. But when you dig into the basics, it becomes clear that there are companies that have used that "new" technology for years already.
Consider the cloud, or desktop as a service (DaaS). Before cloud computing was a trend, there were companies that had already built what we'd now call a private cloud in their data centers. And DaaS was a hot topic in the virtualization market in recent history, but it's just hosting virtual desktops in the cloud, which isn't new either.
What is hyper-converged computing, really? It's the integration of hardware you probably already have on premises using software you probably don't. In that way, it's totally possible -- in fact maybe even easy, in some situations -- to build your own hyper-converged infrastructure with your same old data center components. All you need is the software.
That said, there are pros and cons to the DIY approach to hyper-converged computing that you'll need to know about before you buy software to define your hardware. You might have to buy new infrastructure components if what you have laying around is too old or your chosen software doesn't support it.
Plus, you'll have to research which software is right for your company and ultimately determine if you have the in-house expertise to manage bespoke hyper-converged computing. The benefit of buying a hyper-converged computing appliance is that you can contact one vendor for all your support needs; using disparate hardware and software means you could be looking down the barrel of a support nightmare if something goes wrong.
Whether you choose to build HCI yourself or buy it from a vendor, you won't be in uncharted territory.