There are over a billion PCs in use today that are more than three years old (according to Intel, I've not personally counted them, but I assume they have their finger on the pulse), and around half of those it is claimed are four to five years old. While that sounds like a massive opportunity for the OEMs to cash in on an upgrade wave, the truth is that there's little reason for people to spend the money to replace a working PC.
There's a dirty little secret that PC makers don't want you to know, and it's this - take a middle-of-the-road PC from three years ago and put it next to a brand new middle-of-the-road PC and you'll be hard-pressed to see any difference.
Sure, on paper it's going to be advertised as being faster, and benchmark tests will support this, but in the real world - assuming that the old one isn't all kludged up and ailing - you're just not going to see much difference. Sure, there might be a slightly faster boot up time, or it might be a little snappier, but most of the gains that new PC owners see are nothing more than confirmation bias. Sure, if you go to the performance end of the scale then things are different, but you average home or office PC spends most of their time running a web browser or a word processor application.
A three-year-old PC can do that without breaking a sweat.
And right there is the problem facing the PC industry. You're replacing a tool with another tool that does the same thing. Much like a light bulb or a hammer. It's why we're seeing a proliferation of "smart" devices - smart light bulbs, smart thermostats, smart smoke detectors, smart refrigerators - because without that new "smart" twist people just aren't replacing their light bulbs, thermostats, smoke detectors, or refrigerators until the day they release the magic smoke and stop working.
Modern devices have a good lifespan, and this presents a challenge to anyone trying to sell, and make a decent profit, from these devices.
And this is exactly what's happened in the PC space. Performance (and for that matter other variables such as storage) has plateaued, and there's now little left to drive sales other than there are still people who like shiny new stuff. But even those people would rather spend their money on shiny new smartphones and tablets.
You can see just how deep a hole the OEMs are in, and why they're desperately experimenting with new form factors and such. The OEMs themselves know that there's little value in replacing a functioning PC with a new one, so they're trying to get us to buy new types of PCs. At the end of the day, splurging on an all-in-one or a 2-in-1 system doesn't really change the fact that at the end of the day it's still just a PC, and it's going to be doing PC things.
And those of you paying attention will already know that the "OEMs experimenting with new form factors" tune is nothing new. Remember the netbook, or the home theater PC?
Then throw into the equation the fact that you can load Windows 10 onto almost any system running Windows 7 or 8.x, and that the operating system is free to pretty much anyone who wants it, and all of a sudden the OEMs can't even rely on a new version of Windows to sell new systems. In fact, it's worse than that because the long life that Windows XP enjoyed, combined with the modest system requirements the operating systems that have followed it have had has all helped to highlight to consumers and enterprise alike just how long a PC can now last.
I'd now go as far as to say that the biggest limiting factor in a PC is the hard drive (after about five years of operation, even the healthiest hard drive is on borrowed time), and that as more and more system move to solid state, PCs are going to last even longer.
The future does not look good for PC sales. Sure, people are going to want to buy PCs for the foreseeable future, but booming sales and predictable upgrade cycles are a thing of the past.