CISC vs. RISC
Historically, processors are divided into two camps: one headed by Intel and their clones and one based on Motorola’s technology, backed by companies such as Apple and Sun Microsystems. These camps hold distinctive design strategies to allow them to produce the “fastest” machines while maintaining cost effectiveness. The crux of the debate has been balancing the expense of system programming with the expense of hardware to reduce overall system cost.

Let’s put the technology into perspective. In the 70’s and 80’s, when these camps were forming, hardware was a premium. Storage and memory were expensive, though it was believed prices would drop as technology matured. Writing code for this hardware was high level and fairly complex, making development time extensive and increasingly expensive as technology became more complicated.

The Intel camp determined that to optimize performance, they needed to move the complexity away from software and into hardware. They developed a set of common functions that programmers often use and imbedded them into the core logic of their processors. This allowed the hardware to break down the predefined sets of functions and programmers to call these complex instructions with a single command. By making programming easier, coding took less time and kept overall costs down. This was the focus of processors based on CISC (Complex Instruction Set Computer) architecture.

The Apple camp found that compiling CISC instructions required a lot of hardware support, and that few programmers were actually using all the imbedded complex instructions. Instead, programmers were using smaller, simpler instructions to perform these complex functions. The RISC (Reduced Instruction Set Computer) architecture is based on providing a smaller set of imbedded functions in the hardware (and thus, less hardware), but optimizing the way that code is compiled. By using simpler code that can be compiled more efficiently, optimization comes through software rather than hardware.

Modern Technology
Today’s processors don’t technically belong entirely to one camp or the other.  Intel, AMD, and Motorola are all taking advantage of falling hardware costs. They are consistently shrinking the manufacturing process, making their chips run more efficiently, while adding more transistors to enable more complex operations to be handled by each chip. All are increasing the amount of memory included on their processors (known as cache memory), enabling faster handling of data between the processor and main system memory (RAM). These processors are being combined with the fastest system buses and beefier peripheral components to provide the most robust systems to date.

And while each processor maintains core architectural differences, RISC and CISC computers are moving closer together, with instruction sets becoming more similar and equally complex. The CISC camp became divided with the introduction of the Pentium with MMX (multimedia extensions) from Intel and the K6 with 3D-Now! from AMD. Each introduced their own proprietary multimedia extensions into the core instruction set. Apple responded with the Velocity Engine, its own set of multimedia extensions. Each of these extensions is based on SIMD (Single Instruction, Multiple Data) technology, which allows a certain function to perform a single operation on several pieces of data at once. For instance, an entire group of numbers can be added together with a single SUM (X,Y,Z) command, instead of an ADD command being called again and again to produce the end total of X+Y+Z. Since the processor does not have to recompile a new command for each piece of data, overall processing efficiency is increased. However, to take advantage of these extensions, software must be written that explicitly uses them.

End the debate
The debate on which technology is better is nearly moot at this point. It is far more a matter of marketing, or of opinion based on platform preference, than actually based on any true benchmarks. A trip to nearly any vendor’s Website will reveal tests performed that demonstrate the various strengths and weaknesses of each respective technology. When placing any of these processors in an otherwise equal environment, the top of the line Motorola will handle itself just as admirably as the top of the line Intel or AMD. 

In fact, it is far more a matter of optimizing the programs running on each individual platform. As mentioned earlier, code must be written to take advantage of the advances in technology that each processor takes. If software is optimized to utilize Intel’s extended instruction set, it will perform better on Intel. If the software is optimized to Apple’s technology, then the G4-based systems will provide superior performance.

So which technology is better? It truly depends on what software you run.