This is the 25th day of my participation in the Gwen Challenge.More article challenges

Chapter one: Basic concepts of computer system architecture

Noun explanation:

1.1 VM: A vm implemented by software

1.2 System Structure:

Narrow: The properties of a computer (conceptual structure and functional characteristics) that programmers can see

Generalized: instruction set structure, composition, and hardware

1.3 System acceleration ratio: –

$$Acceleration ratio = System performance (improved)/System performance (before improvement) = Total execution time (before improvement)/Total execution time (after improvement)$$

Note: acceleration ratio > 1

1.4Amdahl’s Law: The system performance acceleration ratio achieved by accelerating the execution speed of a certain component is limited by the percentage of the execution time of this component in the total execution time of the system

1.5 Series computers: a series of different models of computers produced by the same manufacturer with the same system structure, different components and implementations

1.6 Simulation: implement the instruction set of another computer on an existing computer by software method

**1.7 Simulation: ** Interprets the instruction set implementing another computer with a microprogram of an existing computer

**1.8 Parallelism: ** Multiple operations or operations performed by a computer system at the same time or at the same time interval

Short-answer questions:

1. What is software compatibility? How many kinds of software are compatible? Which of these is the essential characteristic of software compatibility?

:curly_loop:** Software compatibility :** A piece of software can be ported from one computer to another with little or no modifications. The difference is just the execution time.

Software compatibility features:

• Upward compatible
• Backwards compatible

Up and down for the computer level

• Forward compatible
• Backward compatibility

Before and after is time

Backward compatibility is the essential characteristic of software compatibility

2. Take a series of computers as an example to illustrate the relationship between computer system structure, computer composition and computer implementation.

1. Computer composition: a logical implementation of the structure of a computer system

2. Computer implementation: A physical implementation composed of computers

3. Computer system architecture can have many components, and one component can have many implementations.

   Each type of machine in the same series has the same system structure, using different composition and implementation, so it has different performance and price

3. What are the three basic principles most often used in the structural design and analysis of computer systems?

• High probability event priority principle
• Amdahl's law
• Principle of program locality

4. According to Amdahl’s law, which two factors determine the system acceleration ratio?

• Component acceleration ratio
• Modifiable ratio

5. What are the three technical approaches to improving parallelism in computer systems?

• Time overlap: Stagger multiple processes in time, alternately overlapping parts of the same set of hardware
• Resource overlap: Dramatically improves the performance of a computer system by repeatedly setting up hardware resources
• Resource sharingA software method that allows multiple tasks to alternate using the same set of hardware devices in a certain chronological order

Supplementary test points:

1. Common classification methods of computer system structure include: Feng’s classification and Flynn’s classification

** Feng classification: ** Computers are classified by the maximum parallelism of the system

**Flunn taxonomy: ** Split according to the polyploidy of instruction stream and data stream

• Single instruction stream single data stream
• Single instruction stream multiple data stream
• Multiple instruction stream multiple data stream
• Multiple instruction stream multiple data stream

2. Quantitative Principles of Computer System Design (4 pieces)

• Focus on recurring events
• Amdahl’s law

$Total execution time _{improved}= time of modifiable parts + time of non-modifiable parts$

$After the modification of the total execution time _ {} = \ frac {can improve ratio * total execution time _ {before improvement}} {components to accelerate than} + (1 – can improve ratio) * total execution time _ {} before improvement$

So:

• CPU performance formula

$CPU time = number of clock cycles required to execute the program * clock cycle time$

$CPI= number of clock cycles required to execute the program/number of instructions executed (CPI: average number of clock cycles per instruction)$

$=> CPU time =IC*CPI* clock cycle time (IC: number of instructions executed)$

• The locality principle of programs

4. Execution time and throughput

• Execution time: The time it takes for a single program to complete
• Throughput: the number of programs executed per unit of time

Chapter two: computer instruction set structure

Noun explanation:

2.1CISC: Complex instruction set computer

2.1RISC: Reduced instruction set computer

2.3 Addressing: How does an instruction set structure determine the address of the data to be accessed

2.4 Data representation: data types that can be directly recognized by computer hardware and directly called by instruction sets

Short-answer questions:

1. What are the disadvantages of CISC architecture from the point of view of current computer technology?

• The frequency of use of various instructions varies greatly
• The instruction set is huge, the instruction number is very many, the function of many instructions is very complex
• Many instructions have large CPI values and slow execution due to complex operations
• Because the instruction function is complex and the regularity is not good, pipeline technology is not conducive to improve the performance

2. What are the design principles of RISC?

• The number of instructions is small and simple
• Adopt simple and unified instruction format
• The execution of the instructions is completed in a single machine cycle
• Only load and store instructions can access the memory. All other instructions operate between registers, namely load-store structure
• Most instructions are implemented using hardwired logic
• Emphasizes the role of the optimization compiler, which generates optimized code for high-level language programs
• Take full advantage of pipeline technology to improve performance

3. Which addressing methods are used by MIPS? (5 or 2, that’s 2 representations)

• Immediate digital addressing
• Offset addressing

Supplementary test points:

1. Classify instruction set structures according to the types of storage units storing operands, and the characteristics of each class (Figure 2.1 and Table 2.1)

• The stack structure
• Accumulator structure
• Universal register structure
  • Register-memory structure
  • Register-register structure

2. Two representations of operand types

• The integer
• Floating point Numbers

3, MIPS three instruction formats

• Class I instruction: including load, store, immediate count instruction, branch instruction, register jump instruction, register link jump instruction
• R class instruction: ALU instruction, special register read/write instruction, move instruction
• Class J instruction: jump instruction, jump and link instruction, trap instruction and abnormal return instruction

4, instruction set three encoding formats

• Variable length encoding format
• Fixed length encoding format
• Hybrid encoding format

5. Four operation types of MIPS

• The load and store
• ALU operations
• Branches and jumps
• Floating point operations