Code compression and code compaction become hot technologies as programmers try to squeeze code size.
Stronger, higher, faster. Smaller? That seems to be the battle cry these days as embedded developers and microprocessor makers strive with Olympian effort to make programs smaller. A concept clearly alien to the developers of Microsoft Office, code density has become a hot topic in embedded circles. There are good reasons why, and lots of ways to accomplish the feat.
What few people predicted was just how the design of a microprocessor can help. Just within the past year or so, no fewer than three different CPU makers have shown off the results of major efforts to improve code density and shrink program size for their chips. Two of them—MIPS and ARM—have taken roughly similar approaches, although there are differences. The third competitor, IBM, has found a radically different and typically exotic solution, delivering perhaps better results but with its own set of trade-offs.
Why bother? What’s all this fuss about code density and why does the choice of microprocessor matter? Read More