The leading companies like, Intel, Motorola and Texas Instruments etc, have contributed significantly towards the growth of mobile processors. The paper compares and critically examines three different products (embedded microprocessors) on the basis of five comparison criteria; power consumption, code density, peripheral integration and chipsets, multimedia accelerators and performance.
Microprocessors came into existence with the advent of computer technology. Since then, there has been a steep increase in the technological advancement with each new technology replacing the old counterpart in a matter of a year or two. Power consumption and dissipation, among these processors, was not a cause of concern as the issue could be easily addressed by use of fans with processors to keep them cool apart from providing an air conditioned environment. This arrangement sounds good for desktop applications only, while mobile applications warrants good performance in open environment with stored power sources (batteries) and no cooling mechanism (Schlett, Manfred., 1998).
With the wireless technology boom and rising demands of services on mobile platforms like, PDAs, cellular and smartphones, the embedded processors faced a number of challenges, of which power management (for longer battery life), processing speed (for better access to services) and memory management became far to important (Conte, Thomas M., 1997). In the subsequent paragraphs, the paper would discuss the basic criteria for judging the overall performance of mobile embedded processors. Further, the paper would critically examine three brands of embedded processors from different companies and compare them.
Criteria for Comparing Microprocessors
Embedded microprocessors have a number of associated properties and features of which the following five offer best criteria for differentiating and estimating the performance:-
Power consumption of an embedded processor, used for mobile applications, is generally done with an established benchmark for comparison - MIPS/watt (Schlett, Manfred., 1998). The power consumption is noted for three different modes of operation; active (when the clock signal traverses to every part of the processor), Standby (when only a few instructions are processed , memory refreshing is affected and goes to active mode on receiving an external interrupt) and power off (the clock is switched off).
Code density refers to the complexity and density owing to code length (Schlett, Manfred., 1998). CISC processors exhibit better code density than RISC processors due to complex instruction set in CISC, while RISC utilizes a fixed length instruction code. However the RISC architecture offers simple and fast instruction decoding.
It is not always advisable to integrate a large number of peripherals and chipsets with the processor. This would increase complexity and may not lead to desired results. It may further, create power consumption problems. A possible way out could be a separate chipset for applications.
Multimedia accelerators refer to the application software that are an important need of the day with demands of mobile net access, video, camera, gaming and