The day before yesterday, I was asked and (after a pause and some thought ), I did an impromptu explanation to few investors on the following subject: “What makes Data Centre (DC) a good investment”. And (lo and behold) I managed to explain one highly technical and potentially highly expensive concept to few real (v.s. potential) investors. And they went away satisfied. Which is always a good thing.
Each and every Data Center has two parts:
- Computers (Hardware + Software)
- Everything else
Where Computers (Hardware + Software) are a well-known quantity in the DC Architecture. To design and implement a “bunch of servers” to act together as a good Data Centre, one more or less “just” needs a good engineer (or a few more of them).
These days we can claim: In the context of DC building. Computers are easy. The market is enjoying the abundance of cheap and excellent hardware. Prices never fail to amaze with the new low’s they are reaching every day. The same is with software. Besides the fact that software never fails to amaze with new heights in license costs. Put all together, one can very successfully predict and calculate the time and money required to assemble the hardware and software for one nice and large data centre. One can even purchase a ready-made “Data Centre in a Container and solve the whole “computers” part in one single step. But.
The key question in the DC building art: Where to put it?
This is where “everything else”, a.k.a. “the second part” comes into the investor’s drama. “Where” issue is the “dark matter” in the DC’s “universe”. And, this diagram was the core of my ad-hoc “presentation”. (with a layer of readable text added later by me). What it shows is how to qualify the answer to this key issue, with the little help of one whiteboard drawing.
Imagine a location that is offered to you to house your future investment. Your multi-million dollar data centre. Will you take it? Will you invest in it? How will you actually decide?
To be feasible, to be used as the data centre that physical location has to be in the logical sweet spots region, defined and presented on the diagram.
There are three key attributes to be taken into account when qualifying each data centre:
Key Attributes: Energy, Technology and Security
Or in slightly more detail:
- Energy: data centre needs a lot of it, and one needs it to be as cheap as possible.
- Technology: data centre needs to contain a sizeable amount of quality workforce. Also, there has to be quality infrastructure taking signals in/and out of the data centre.
- Security: data centre needs to be in a location that is financially secured, in a politically stable country and in the immediate (and wider) environment which is friendly towards the data centre, its owner and its purpose.
And here we have, at the same time, the key requirements to be solved to build one data centre one can call (later) feasible. Computers are easy. Your future DC has to be in the “region of sweet spots” between the three key attributes.
I think this is one succinct explanation on what makes a feasible date centre, and therefore on the subject of what are the key requirements to make it a good investment.
Enjoy … and re-use :)