Confessions of a Serial Mapper

There seems to be a lot of chatter currently going around hardware forums that for the first time in history, computing power has outstripped the pace of computer games.  Meaning that games are being able to be run on new computers much easier than they may have been in the past.  This is a result of the number of cores of CPU’s increasing rapidly as recently the new 6-core Intel processor is set to be released later this year.  So what do all these additional cores actually mean?

Well, by having extra cores, you can technically set one core to be used for physics, one to handle sound, one to handle general computations and one to handle the AI.  So we can expect developers to start creating games that will take advantage of these additional cores fully, leading to better games more rapidly, as double the number of cores would logically lead to double the computing power available, right?  Well, sort of, whilst an increase in cores does mean that there is more total processing power inside of a computer, unless it can be tapped into by developers, it is useless.

Let me give you an example, the release of the Orange Box from Valve was meant to bring in a new version of the Source game engine that would allow the engine to use multiple CPU cores.  However, this is not fully implemented and took Valve a lot longer to perfect, in fact I am sure they are still trying to iron out all of the bugs.  Whilst the games can use 2 core and 4 core processors, the benefits from using the additional cores have yet to be seen.  Multiple cores are an exciting challenge for game developers and is in fact useful for the actual development of games.  The build programs that the Source engine used (VBSP, VRAD and VVIS) can all operate on multiple cores, severely reducing the time it takes to compile a map, very useful for developers like myself.  These benefits include better AI, superior particle physics and more realistic physics.  These will not be seen until Episode 3, when the game is programmed with multi-cores in mind.

I am wondering if there will be a point that computers will never need to get any more powerful for the purpose of gaming, we are already entering the age of photo-realistic games, they look very close to real life.  However, there is still some obvious, more technical issues that still tell us it is only a game.  Firstly, getting accurate physics is still an issue and having a fully interactive environment (more so than Crysis) is still a long way off.

There is also the cost of trying to make games look and feel better and better.  By trying to use up all of this additional power, by having incredible graphics, realistic physics, smart AI etc.  It pushes development costs up and up, which will in turn either effect the companies profits, or more likely, increase the cost of the product for us, the gamers.  It will also push away modders from working with the games engine, as it’s complexity increases, so does the knowledge needed to use it to it’s fullest does.  Instead, game developers need to take incremental steps, as to prevent a sudden sky rocket in costs, at roughly the same rate as they always have.  So don’t expect to see games get more rapidly better with the rapidly increasing technology, but instead expect to be able to play newer games to their fullest with a less than top-end computer.