Designing Embedded Hardware, 2nd Edition

Chapter 17. MAXQ

We will find a way, or we will make one

—Hannibal, 218 B.C.

In this chapter, we'll look at an innovative new processor architecture introduced to the world in 2004. Dallas Semiconductor, a subsidiary of Maxim (http://www.maxim-ic.com), developed the 16-bit MAXQ microcontrollers to target the low-cost, low-power embedded applications market. The architecture is aimed directly against Microchip's PIC, Atmel's AVR, Texas Instruments' MSP430, and the 8051 architecture offered by many manufacturers (including Dallas Semiconductor itself). The MAXQ is an interesting contender for top RISC microcontroller. It's fast, has a lot of functionality, and is very low-powered. At the time of writing, the User's Guide for the MAXQ is 230 pages long. Obviously, this processor has a lot of features, too many to be thoroughly covered here. Therefore, I'm going to simply concentrate on the basic design for a MAXQ-based system. Let's start by seeing what makes the MAXQ so different and so interesting.

Architectural Overview

The stated design goal for the MAXQ was to achieve a high performance-to-power ratio. In other words, the aim was to maximize the processor's throughput of instructions while minimizing the current draw. Many RISC processors achieve single-cycle execution but do so through the use of an instruction pipeline. In a pipelined architecture, the execution unit is comprised of many stages. At any one time, several instructions will be in the process of being decoded ...