Shortly after the release of the Nintendo 64, I rented one and brought it home. I told anyone that would listen that this dark grey, swooped rectangle represented the most powerful graphics processing machine that you could sensibly buy for your living room. There were PC graphics cards, and specialty high end workstation components, of course, that would run shaded polygonal rings around the N64, but they ranged in price from several hundred to several thousand dollars. Here was a machine that was purpose built for drawing triangles to a screen as fast as possible, and it cost around $200.
The N64 was hamstrung in pretty much every other way a computer can be. Not enough memory, outdated software storage, and proprietary everything. Since Nintendo had control of the machine from top to bottom, they could make it do things that seemed to sidestep those limitations. The dismal quality of many third party games proved that maybe Nintendo didn’t really like to share it’s secrets. It hurt the system, and kneecapped the spread of that particular hardware, capable as it was, to other computers and application.
Things in the game industry were changing from hardware centric software design, that tried to squeeze the most performance out of the actual chips, to API and driver centric design. This meant that game developers (or any software developer really) would write their game to talk to an intermediate layer that sits between their application and the actual hardware. A game developer could be less and less concerned with how a specific graphics or CPU chip worked, and focus on OpenGL or DirectX. Those API’s would then in turn be expected to deal with the wide variety of graphics and sound chips available. OpenGL and DirectX were nascent products and they weren’t always reliable, but software is not static like hardware is, so massive improvements came quickly. It wasn’t long before using a particular graphics card and a modified driver let a pretty standard PC emulate the Nintendo 64. The industry had moved on, and it wasn’t the hardware, but the software that pushed everything forward.
The current generation of consoles, namely the Xbox One and the Playstation 4, are headed toward mid cycle hardware upgrades. It will be a new, up to date, version of the same console at the same, or similar, price point. Just faster. Typically a change like this in a console would warrant a new name and some new gimmick to let consumers know that this console is worth the upgrade.
Internally, the Xbox One and the Playstation 4 are PCs. Purpose built versions of off the shelf hardware fit for a low to mid tier gaming computer. They can’t compete with the top of the line Nvidia or AMD PC chips, but they are by no means underpowered. One high end Nvidia card also costs twice what one of the consoles costs, and that is before you even put the rest of the computer around it. I have read a lot of comments that the consoles are lagging behind, but price to performance, they are extremely impressive. More importantly, the hardware doesn’t really matter.
The PS4 runs a Unix type OS and uses an API layer that, by all accounts, closely resembles DirectX. The Xbox One does use DirectX. Games, in almost all cases, aren’t written for the hardware anymore. They are written for the API that talks to the hardware. The hardware could change and as long as the API is aware of that, the games will continue to run. Probably much like they ran on other hardware.
As soon as the announcement about console manufacturers opting for a mid-cycle upgrade broke, I saw several media outlets and commenters claiming that this was the end of consoles. Different versions of a console would stratify the market and create confusion among consumers. You would buy a newer game thinking that it would work in your older system and the whole house of cards would be swept away in a gust of specs and frequencies. Making these consoles into upgradable PCs will ruin the industry.
The computer I’m typing this on is roughly 6 years old, but with a 2 year old video card. The PC sitting 6 feet away from me is almost ten years old. You can play the same games on both. One does a bit better on some games, but neither is any slouch. I have played very new games on the 6 year old PC, and aside from the loss of some visual gloss, they play fine. One uses Nvidia and Intel, the other AMD, but they games don’t seem to care much. There is a PC in the basement older than both of them put together, and I managed to spin up linux and a fairly playable copy of minecraft on the thing a while ago.
The hardware doesn’t really matter any more. Game developers will continue to develop for the average. The computer or console that the most people own will be the development target. A newer Playstation with better chips will offer some headroom, but until the majority of people own that version no game will be made for specifically for that new hardware spec. More importantly, Sony can buy whatever the best price to performance chips are, at commodity prices, and not have to keep a factory busy producing older chips to support a single product. They can keep making the PS4 for 15 years, and it won’t really matter what chips are in it, because your games will probably run. This isn’t like the N64. If it ever gets to a point where new games don’t look as good as you like, or they are getting too slow to be fun, Sony will be more than happy to sell you the newest version for the same $400.
So yeah, consoles have become like PCs, and it’s about damn time.