Actually, it depends. Modern personal computers use mainly CISC (complex instruction set) as oppsoed to RISC. Back in the days of the Motorola 68XXX in the old Macintosh (Apple Computer) days, they typically used a RISC setup until the end of the Power PC era, when of course they moved to Intel.
Mainly the strong points of any processor are that it needs to handle the data thrown at it, and it needs to process it with as few errors as possible. Computer processors themselves are a different breed than most audio I/C's in that they are designed to predict when and what to draw from system memory, and if the processor fails to acquire the correct area of memory, then it has to clear the stack and re-try the calculation. It equates to the CPU being correct, but at the expense of time (given nanoseconds or more.)
An audio processor I'm very sure is set up a much different way. One way such that it's source of data to be processed is true and correct in value and buffer, and that the processor is designed for continuous frame processing. The key to a audio processor is simplicity, in it's basic form. Making an audio chip too complicated can lead to many features and advancements, but possibly at the cost of a company dealing with a very steep learning curve to utilize a chip that is more complex, the addition of not being able to use certain features efficiently or effectively, or the added cost of a product that may do the same or less than a comparable model of another brand.
