(Part 2 in a series of posts on small-form-factor computing)
Today’s PCs and DIY desktops come in a dizzying array of sizes:
We are familiar with the idea that any device can be “scaled up”, made bigger in size: our portable media device-things come in sizes ranging from 4″ to 10″. But what are they like inside? If you make them bigger, do the parts stay the same while the amount of empty space increases?
I like looking at teardown videos/pages, just to see what actually goes in there, how small/large the various components are, but more importantly how much empty space there is. To me, empty space represents lots of opportunity for optimisation: is it serving a purpose? Can its purpose be served by something else that uses less space?
Smartphones and tablets represent the apex of this reduction in empty space: every part of that chassis has to be put to use, if you want your device to be as thin as possible.
The interesting thing about smartphones and tablets is, beneath the surface they all look much the same. Surface impressions of size and colour aside, once the back cover is off, we see much the same thing. Battery. Circuit board and various connectors. Screen underneath.
When you scale up an iPhone to an iPad, what happens to the various parts? The battery and screen definitely increase in size; what about the circuit board?
The iPhone circuit board is smaller, because some of the components are on a separate board. But look at the size of those things: the chips aren’t much bigger, or smaller. Why would we expect otherwise? Guess what: it’s the same thing with desktops.
Desktop scaling #
So what’s actually happening when we shrink a desktop? There’s no battery, no screen to think or worry about. How do those components inside change to fit their new skin?
Exhibit A: Corsair 750D (full tower) #
Let’s get the obvious out of the way: more than half of that case is just empty space. This is useful if you’re running a high-end, overlocked system with multiple graphics cards, since lots of air usually means good ventilation and thus better cooling. But not all of us fit in that category; I don’t have three hard drives and two graphics cards in my desktop! What about those of us who want something smaller? How much can we reduce that empty space by?
Exhibit B: Silverstone TJ08-E (mini tower) #
Same components (minus one graphics card, although the TJ08-E is capable of fitting two). By this I do not mean to suggest that the TJ08-E is in the same caliber as the 750D. Our point here is to show that if we don’t have extreme computing needs (and if we’re honest with ourselves, few of us do), reducing empty space with the same components is definitely possible.
One step further? I’m going to do away with the empty-case image, I think we’ve had quite enough of that.
Exhibit C: Bitfenix Prodigy M #
Whoa, now we’re getting cosy. There’s no optical drive labelled there, even though there clearly is a slot, because using it would mean we can’t fully use that upper space for cooling. And now we’re down to one graphics card as well. And once we’ve shrunk things this much, there simply isn’t enough standing space for full hard drive cages as well—they will have to sit on the floor.
But it’s doable.
I have an apology to make. So far, we have not crossed any thresholds just yet; we’ve only been shrinking empty space. Next post, I promise.
Can we go further than this? Clearly some compromises, comfortable or uncomfortable, will have to be made. But let’s not throw in the towel yet, let’s push on and see what these demands are, and what thresholds they might enable us to cross.
Wait, what about cooling performance?! #
As far as I know, these three desktop cases have fared well in cooling performance in various reviews. As one goes smaller in case size, there will be some decrease in cooling performance. But as we shall see in later posts, going to certain sizes gives us a bag of tricks that may actually lead to better cooling performance over some medium-sized cases.