The Hard Science of Making Video Games

Like re-creating the ceiling of the Sistine Chapel with a couple magic markers

Problem: If a computer can´t keep up with the instructions the game issues, the image stutters, ruining the experience. That threshold is the number-one headache of making games. It compounds every challenge listed here, from building artificial intelligence to obeying physics.

Status: Although games now take advantage of â€multi-core processingâ€-the simultaneous use of multiple central-processing or graphics-processing units-programmers are always seeing their fantasies throttled by the need to budget processing capacity. (The PlayStation 3 contains eight 3.2-gigahertz processors, but few designers yet know how to program for them.) The budgetary mindset is so entrenched, programmers use the language of accountants (â€ugh, this rain´s too expensiveâ€) to describe the problem.

What´s next: Moore´s law-which states that the number of transistors on a chip doubles roughly every two years-means that there will always be more processing power available. (Nvidia, a leading manufacturer of graphics chips, claims it has managed to do Moore one better, doubling its chips´ processing power in less than a year.) But programmers´ ambitions will always outstrip the hardware because, as one designer puts it, â€the more we can do, the more excited we get, and the more we want to do.†With every jump in CPU and GPU capacity, that frustration only mounts.-Jacob Ward

Image: The PS3's Cell processor