7. Faster and Faster

AMD countered the Intel Pentium II MMX with their 3Dnow! SIMD design. This expanded the SIMD concept with floating point calculations that extended the range of numbers that can be crunched. Only a year later, the Intel Pentium III came out with SSE, or Streaming SIMD Extensions that further improved on the MMX design with more flexibility with registers expanding to 128-bits.

Note that SIMD is not multi-threaded or even multitasking. SIMD only works when special instructions are used by the programmer. The data has to be organized in a very strict format to fit into the SIMD registers. Even with all the restrictions, SIMD freed the PC as a graphics processing powerhouse that unseated the specialized graphics silicon from companies like Silicon Graphics.

8. Contrary to KISS

In most cases, KISS (Keep It Simple Stupid) works. Simple is better than complicated, except in the case of microprocessors. One group of companies tried to keep to the KISS principle and designed Reduced Instruction Set Computer (RISC) microprocessors. They believed that by keeping the instructions very simple they could make the microprocessor perform faster that it would make up for having to use more of the simple instructions.

Intel went the CISC (Complex Instruction Set Computer) route where they have many many instructions and lots of those could do some pretty complicated things. What the RISC guys didn't count on was that Intel and AMD could fabricate their own microprocessors with the complexity of more than 100-million transistors and get that complex machine running at gigahertz rates.

Scientists and engineers had favored the expensive RISC workstations for their believed superior speed performance over the PC. When it became obvious that RISC was a dead end, scientific and engineering software migrated to the PC with those inexpensive yet powerful CISC microprocessors.

9. Supercomputers On the Cheap

The supercomputer is the ultimate number crunching machine for physicists to work out the mysteries of the universe, meteorologists to predict the weather or for doctors to dissect human genes. They have to process lots of numbers over and over. Hey, wait a minute, that's just like the video processing for our video games. Well, guess what. Those supercomputers are built out of the same microprocessors that are the brains of our PCs! They just connect lots of them together to break the problem up into smaller chunks and process the data in parallel.

Final Words

That was some pretty dense technical stuff! I hope you caught the gist of it because we'll be using a lot of these terms in the next installment. All this talk about processing data in parallel is leading up to something. We'll get around to talking about hyper-threading and how even within a single microprocessor there are a lot of things going on in parallel.

It will lead to other features like superscalar processing where more than one instruction can be executed during a single clock cycle. All this to make your programs run faster!

Tech Tips Article - http://www.geeks.com/techtips/