At present, the high-end products of embedded processors are: Advanced RISC Machines ARM, Silicon Graphics MIPS, IBM and Motorola's Power PC, Intel's X86 and i960 chips, AMD's Am386EM, Hitachi's SH RISC, Harvatek chip.

Microprocessors and microcontrollers

        Microprocessors and microcontrollers of the difference, this difference is mainly concentrated in the hardware structure, application areas and instruction set features three aspects:

        一、the hardware structure. Microprocessor is a single-chip CPU, while the microcontroller in an integrated circuit chip integrates CPU and other circuits, constitute a complete micro-computer system. In addition to the CPU, the microcontroller also includes RAM, ROM, a serial interface, a parallel interface, a timer and an interrupt scheduling circuit. Although the on-chip RAM capacity is smaller than that of an ordinary microcomputer system, this does not limit the use of microcontrollers. In the later can understand that the scope of application of a wide range of microcontrollers. Among them, the microcontroller is an important feature of the built-in interrupt system. As a device for control, microcontrollers often respond to external incentives in real time.

        二、the application areas. Microprocessors are typically used as CPUs in microcomputer systems, and their design is precisely for such applications, which is also the advantage of microprocessors. However, microcontrollers are typically used for control-oriented applications, and system designs are designed to be miniaturized to minimize the number of components. In the past, these applications usually need to use dozens or even hundreds of digital integrated circuits to achieve. The use of microcontrollers can reduce the number of components used, only a microcontroller, a small amount of external components and stored in the ROM control program can achieve the same function. Microcontrollers are suitable for applications where input / output devices are controlled with very few components, and microprocessors are suitable for information processing in computer systems.

        三、the instruction set features. Due to the different applications, microcontrollers and microcontrollers are also different instruction sets. The microprocessor's instruction set enhances the processing function to have a powerful addressing mode and instructions for manipulating large-scale data. The microprocessor's instructions can operate on nibbles, bytes, words, and even double words. By using address pointers and address offsets, the microprocessor provides an addressing mode that can access large amounts of data. The self-increasing and self-decreasing modes make it very easy to access data in bytes, words or double words. In addition, the microprocessor also has other features, such as user programs can not use privileged instructions.

        The instruction set of the microcontroller is suitable for input / output control. Many input / output interfaces are single / bit. For example, the electromagnet controls the motor's switch, and the electromagnet is controlled by a 1-bit output port. The microcontroller has instructions for setting and clearing units, and can perform other bit-oriented operations, such as logical AND, OR and exclusive-OR operations on bits, jumps based on flag bits. Few microprocessors have these powerful bit manipulation capabilities because designers are only thinking of manipulating data in bytes or larger when designing microprocessors.

DSP chip classification and features

        The CPUs in the digital signal processor are designed specifically for discrete time signal processing calculations, such as those where audio and video communications are required. Particularly suitable for digital signal processing operation of the microprocessor, its main application is real-time fast to achieve a variety of digital signal processing algorithms. According to the requirements of digital signal processing, DSP chip generally has the following main features:

(1) a multiplication and one addition can be performed in one instruction cycle;
(2) the program and the data space separately, can simultaneously access the instruction and the data;
(3) The chip has a fast RAM, usually through a separate data bus in two pieces at the same time access;
(4) hardware support with low overhead or no overhead cycles and jumps;
(5) fast interrupt handling and hardware I / O support;
(6) having a plurality of hardware address generators operating in a single cycle;
(7) multiple operations can be performed in parallel;
(8) to support the pipeline operation, so that fetch, decoding and execution operations can be implemented in overlapping.

         Of course, compared with the general-purpose microprocessor, DSP chip other common functions are relatively weak, DSP chip internal program and data separate Harvard structure, with a dedicated hardware multiplier, extensive use of pipeline operation, to provide special DSP Instruction, can be used to quickly achieve a variety of digital signal processing algorithms. DSP chips can be classified in the following three ways.

1、according to the basic characteristics of points

        This is based on the DSP chip working clock and instruction type to classify. If a clock frequency in any clock frequency, DSP chip can work properly, in addition to computing speed changes, there is no performance degradation, such DSP chip is commonly known as static DSP chip. If there are two or more DSP chips, their instruction set and the corresponding machine code machine pin structure compatible with each other, this kind of DSP chip called the consistency of DSP chip. For example, the United States TI's TMS320C54X belongs to this category.

2、according to the data format points

        This is based on the DSP chip to work the data format to classify. Data DSP chip in fixed-point format is called fixed-point DSP chip, such as TI's TMS320C1X / C2X, TMS320C2XX / C5X, TMS320C54X / C62XX series, ADI's ADSP21XX series, AT & T DSP16 / 16A, Motolora MC56000 and so on. Floating point DSP chip working in floating point format, such as TI's TMS320C3X / C4X / C8X, ADI's ADSP21XXX series, AT & T DSP32 / 32C, Motolora MC96002 and so on.

        Different floating-point DSP chip used in the floating-point format is not exactly the same, some DSP chip using a custom floating-point format, such as TMS320C3X, and some DSP chips are used IEEE standard floating-point format, such as Motorola's MC96002, Fujitsu's MB86232 and ZORAN's ZR35325.

3、according to the use of points

        According to the use of DSP to points, can be divided into general-purpose DSP chip and dedicated DSP chip. Universal DSP chip for ordinary DSP applications, such as TI's series of DSP chips are common DSP chip. Dedicated DSP chip is designed for specific DSP operations, more suitable for special operations, such as digital filtering, convolution and FFT, such as Motorola's DSP56200, Zoran's ZR34881, Inmos IMSA100, etc. are dedicated DSP chip.