Architecture:Types of Microprocessors

The microprocessor industry is divided into the computer and embedded sectors. Both computer and embedded microprocessors share aspects of computer system design, instruction set architecture, organization, and hardware. The term architecture is used to describe these fundamental aspects, and more directly refers to the hardware components in a computer system and the flow of data and control information among them. In this section, various types of microprocessors will be described, fundamental architecture mechanisms relevant in the operation of all microprocessors will be presented, and microprocessor industry trends discussed.

Types of Microprocessors

Computer microprocessors are designed for use as the central processing units (CPUs) of computer systems such as personal computers, workstations, servers, and supercomputers. Although microprocessors started as humble programmable controllers in the early 1970s, virtually all computer systems built after 1990 use microprocessors as their CPUs. The dominating architecture in the computer microprocessor domain today is the Intel 32-bit Architecture, also known as IA-32 or X86. Other high-profile architectures in the computer microprocessor domain include Intel Itanium Processor Family (IPF), HP PA-RISC, SUN Microsystems SPARC, and IBM/Motorola PowerPC.

Embedded microprocessors are increasingly used in consumer and telecommunication products to satisfy the demands for quality and functionality. Major product areas that require embedded microprocessors include digital TV, DVD players, digital camera, network switches, high-speed modems, digital cellular phones, video games, laser printers, and automobiles. Future improvements in energy consumption, fabrication cost, and performance will further enable new applications such as intelligent hearing aid. Many experts expect that embedded microprocessors will form the fastest growing sector of the semiconductor business in the next decade [1].

Embedded microprocessors have been categorized into DSP processors and embedded CPUs due to historic reasons. DSP processors have been designed and marketed as special-purpose devices that are mostly programmed by hand to perform digital signal processing computations. A recent trend in the DSP market is to use compilers to alleviate the need for tedious hand coding in DSP develop- ment. Leading DSP microprocessor vendors include Texas Instruments, Lucent Technologies, Motorola, and ST.

Embedded CPUs are often used to execute operating system, networking, and user-interface code in a consumer and telecommunication products. They have been traditionally derived from out-of-date computer microprocessors. Embedded CPUs often reuse the compiler and related software support developed for their computer cousins. Recycling the microprocessor design and compiler software minimizes engineering cost. Major vendors of embedded CPUs include IBM, Motorola, ARM, and MIPS.

An important recent trend in the embedded microprocessor market is toward integrating an embedded CPU, a DSP processor, and application-specific logic to form a single-chip solution. This approach is enabled by the ever-increasing transistor density achieved by the semiconductor fabrication technology. The major benefit is reduced system cost and energy consumption. In these designs, embedded CPU and DSP processor vendors no longer market microprocessor chips. Rather, they make their designs available for licensing by solution developers such as a cell phone vendor. The embedded CPU and the DSP processors thus incorporated into these single-chip solutions are called embedded CPU cores and DSP processor cores. For example, MIPS customized its embedded CPU core for use in Nitendo64. IBM, ARM, ST, NEC, and Hitachi offer similar products and services. Owing to an increasing need to perform DSP computation in consumer and telecommunication products, an increasing number of embedded CPUs have extensions to enable more effective DSP computation.

There are several ways in which the needs of embedded computing differ from those of more traditional general-purpose computing systems. Constraints on the code size, weight, and power consumption place stringent requirements on embedded microprocessors and the software they execute. Also constraints rooted in real-time requirements are often a significant consideration in many embedded systems. Furthermore, cost is a severe constraint on designing and manufacturing embedded processors.

In spite of the different constraints and product markets, both computer and embedded microprocessors share many main elements in their design. These main elements will be described. Additionally, over the past decade, substantial research has gone into the design of microprocessors embodying parallelism at the instruction level as well as aggressive compiler optimization and analysis techniques for harnessing this opportunity. Exploitation of parallelism is a very effective approach to reducing the power consump- tion under given performance requirements. Much of this effort has since been validated through the proliferation of high-performance general-purpose microprocessors, such as the Intel Itanium II proces- sor, based on these technologies. Nevertheless, growing demand for high performance in embedded computing systems is creating new opportunities to leverage these techniques in application-specific domains. The research of instruction-level parallelism (ILP) has developed distinct architecture meth- odologies referred to as very long instruction word (VLIW) and explicitly parallel instruction computing (EPIC) technology. Overall, these techniques represent fundamental, substantial changes in computer architecture.