[Supplement to the new version of the system architecture] - Embedded Technology

Embedded Microprocessing Architecture

von Neumann structure

Traditional computers use the Von Neumann structure, also known as the Princeton structure, which is a memory structure that combines program instruction memory and data memory

• Von Neumann's computer program and data share a storage space, and the program instruction storage address and data storage address point to different physical locations in the same memory
• Using a single address and data bus , the program instruction and data width are the same
• When the processor executes an instruction, it first fetches the instruction from the memory to decode, and then fetches the operand to perform the operation . Even a single instruction takes several or even dozens of cycles. During high-speed operations, there will be a bottleneck effect on the transmission channel.

Harvard structure

The Harvard structure is a parallel architecture , and its main feature is that the program and data are stored in different storage spaces , that is, the program memory and the data memory are two independent memories, and each memory is independently addressed and accessed independently .

• Corresponding to the two memories are two independent address buses and data buses in the system
• This separation of program and data buses increases execution speed by allowing instruction words (from program memory) and operands (from data memory) to be fetched simultaneously in a single machine cycle , enabling a doubling of data throughput

Microprocessor classification

• According to the word length and width of the embedded microprocessor, it can be divided into 4 bits, 8 bits, 32 bits and 64 bits. Generally, 16-bit and below are called embedded microcontrollers, and 32-bit and above are called embedded microprocessors.

• According to the degree of system integration, it is divided into two categories: one is that the microprocessor contains only a simple central processing unit, which is called a general-purpose microprocessor; the other is that the CPU, ROM, RAM and I/O, etc. components are integrated on the same chip, called a single-chip microcontroller

• If it is divided according to the purpose, it is generally divided into embedded microcontroller MCU, embedded microprocessor MPU, embedded digital signal processor DSP, embedded system on chip SOC, etc.

• The typical representative of the embedded microcontroller MCU is the single-chip microcomputer. Compared with the embedded microprocessor, the biggest feature of the microcontroller is single-chip, the volume is greatly reduced, so that the power consumption and cost are reduced, and the reliability is improved. Its on-chip peripherals Resources are generally abundant, suitable for control, and are the mainstream of the embedded system industry

• The embedded microprocessor MPU is evolved from the CPU in the general computer. It is characterized by a processor with more than 32 bits and high performance. Of course, its price is correspondingly higher, but it is different from the computer processor. In the actual embedded application, only the functional hardware closely related to the embedded application is reserved, and other redundant functional parts are removed, so as to realize the special requirements of the embedded application with the lowest power consumption and resources. Compared with industrial control computers, embedded microprocessors have the advantages of small size, light weight, low cost, and high reliability. At present, ARM, MIPS, POWER PC, etc. are common

• Embedded digital signal processor DSP is a processor specially used for signal processing. It has a special design in terms of system structure and instruction algorithm, and has high compilation efficiency and instruction execution speed. Adopting Harvard structure and pipeline processing, its processing speed is 10-50 times faster than the fastest CPU, and DSP has been widely used in various instruments such as digital filtering, FFT, and spectral analysis.

• Embedded system-on-chip SOC is an integrated device that pursues the maximum tolerance of the product system. The biggest feature of SOC is that it successfully realizes the seamless integration of software and hardware, and directly embeds the code module of the operating system on the processor chip. It is an integrated circuit with a special purpose. , which contains the complete system and has all the contents of the embedded software

multi-core processor

• Multi-core refers to multiple microprocessor cores, which are two or more microprocessors packaged together and integrated into one circuit. A multi-core processor is a single chip that can be directly inserted into a single processor socket. In multi-core technology, scheduling is performed by the operating system software, and multi-process and multi-thread concurrency are all possible.

2 or more core work coordination implementation methods:

• Symmetrical multi-processing technology SMP: two identical processors are packaged in one chip to achieve double or nearly double processing performance and save computing resources
• Asymmetric processing technology AMP: The two processing cores are different from each other, each processing and executing specific functions, and sharing different computing tasks under the coordination of software

Scheduling of multi-core CPU:

• Scheduling algorithms for processes in a multi-core CPU environment generally include global queue scheduling and local queue scheduling. Global
  queue scheduling means that the operating system maintains a global task waiting queue. When a CPU in the system is idle, the operating system waits from the global task. The ready task is selected from the queue to execute, and the CPU core utilization is high.
  Local queue scheduling is that the operating system maintains a local task waiting queue for each CPU core. When a CPU core in the system is idle, it selects an appropriate task from the task waiting queue of the core to execute. switch between CPUs

Soft test-system architect knowledge points extraction-system architect tutorial (2nd edition)

1. Chapter One Introduction
2. Chapter 2 - Basic knowledge of computer systems (1)
3. Chapter 2 - Basic knowledge of computer systems (2)
4. Chapter 3 - Basic knowledge of information systems
5. Chapter 4 - Basic Knowledge of Information Security Technology
6. Chapter 5 - Basic Knowledge of Software Engineering (1)
7. Chapter 5 - Basic knowledge of software engineering (requirements engineering and system analysis and design)
8. Chapter 5 - Basic knowledge of software engineering (software testing, clean room software engineering, component-based software engineering, software project management)
9. Chapter 6 - Basic knowledge of database design (basic concepts of database, relational database)
10. Chapter 6 - Database Design Basics (Database Design)
11. Chapter 7 - Basic knowledge of system architecture design (architecture-based software development method)
12. Chapter 7 - Basic knowledge of system architecture design (architecture style, reuse)
13. Chapter 8 - System Quality Attributes and Architecture Evaluation
14. Chapter 9 - Software Reliability Fundamentals
15. Chapter 10 - Evolution and Maintenance of Software Architecture
16. Chapter Eleven - Future Information Technology
17. Chapter 12 - Information System Architecture Design Theory and Practice
18. Chapter 13 - Hierarchical Architecture Design Theory and Practice
19. Chapter Fourteen - Cloud Native Architecture Design Theory and Practice
20. Chapter 15 - Service-Oriented Architecture Design Theory and Practice
21. Chapter 16 - Embedded System Architecture Design Theory and Practice
22. Chapter 17 - Communication System Architecture Design Theory and Practice
23. Chapter 18 - Security Architecture Design Theory and Practice
24. Chapter 19 - Big Data Architecture Design Theory and Practice