计算机体系结构试题及答案Computer architecture questions and answers.docx

1、计算机体系结构试题及答案Computer architecture questions and answers计算机体系结构试题及答案（Computer architecture questions and answers）Questions and answers of computer architecture1, benefiting from the development of high performance computer: (1) the development of circuit technology; (2) the development of computer ar

2、chitecture technology.2, structure: computer systems can be classified by function of language as a multilevel structure, each layer in different language features. Sixth: the application of virtual machine language - fifth: high-level language virtual machine assembly language - Fourth: - Third: vi

3、rtual machine operating system virtual machine - level second: machine language (traditional machine level) - Level 1: micro machine level program.3, computer architecture: see computer programmer attribute, namely the general structure and functional properties.4, transparency: in computer technolo

4、gy, the things or properties already exists, the concept from a point of view and have called transparency.5, the proposed architecture Amdahl attribute refers to the computer machine language level programmers see.The essence of 3 6, a classic computer architecture concept is to determine the compu

5、ter system hardware and software interface, which is the instruction set design, above the interface by software function realization, interface by hardware and firmware functions to achieve.7, computer organization is the logic of the computer system; computer is a physical computer system to achie

6、ve.The difference and connection between the 8, computer architecture, computer organization, computer?Answer: a system structure can have a variety of components, a component can have a variety of physical implementation, including system structure research on organization and implementation.9, a s

7、eries of machine: refers to a system of the same structure but with different organization and implementation of a series of different types of machines.10, software compatibility: the same software can be run without change on the machine system of the same structure, and the results they get the s

8、ame, the only difference is the different running time.11, compatible machine: different manufacturers, has the same computer architectures.12, backward compatibility is the basic characteristics of software compatibility, is the fundamental characteristics of series machine.13, in the field of comp

9、uter market can be divided into three major areas: servers, desktop systems, embedded computing.14, Moore: integrated circuit density approximately doubled every two years.Technology based analysis of 15 quantitative performance evaluation: (1) (a) response time: from the beginning to the end of the

10、 time between events; all the time spent on the computer to complete a task. (b): the complete flow in unit time and workload. (c) x, y assumed two computers; X faster than y means: for a given task, the response time of X is less than y. The performance of X is several times the Y refers to the res

11、ponse time of X / y = n response time, response time and performance is inversely proportional to.16, the probability of event priority principle: (basic idea) for the probability of events (the most common event), giving priority to use it right and resource rights, to obtain the global optimal res

12、ults.17, Amdahl Law: accelerate the execution speed of a component system performance obtained speedup, the importance is limited by the components in the system of. System acceleration ratio = total execution time (improved) / total execution time (improved) =.18, Amdahl law corollary: if only for

13、a part of the computer in the performance improvement is more improved, the effect of the system. If only a part of the whole task is optimized, so much the acceleration ratio is not greater than 1 / (1- improvement ratio).19, the performance of CPU: Cpu time = total number of clock cycles / clock f

14、requency Cpi = total number of clock cycles (IC / CPI: the number of clock cycles per instruction; the average IC implementation process: the number of instructions.)The performance of Cpu formula: total CPU time = CPI * IC / CPI clock frequency which reflects the computer architecture and computer

15、technology, computer instruction set; Ic reflects the structure and technology of computer programming instruction set; clock frequency: reflect the implementation of computer technology, production technology and computer organization.20, parallelism refers to at the same time or two or more than t

16、wo kinds of properties of the same or different work in the same time interval.The second chapter1, according to the CPU internal storage unit type of instruction set architecture for classification, can be divided into the stack based instruction set architecture, instruction set architecture and a

17、ccumulator type general register type instruction set architecture.2, general register type instruction set machine is further subdivided into 3 types:Register to register type (R-R), register memory type (R-M), a memory register.3, addressing: (1) register addressing: example: ADD R4 R3, meaning Re

18、gsR4-RegsR4+RegsR3(2) immediate values: example: ADD R4, addressing 3 meanings: RegsR4-RegsR4+3(3): offset cases: ADD R4, 100 (R1) meaning: RegsR4+Mem100+RegsR1(4) register indirect addressing: example: ADD R4 (R1) meaning: RegsR4-RegsR4+MemRegsR1(5) index addressing: example: ADD R3 (R1+R2) meaning

19、: RegsR3-RegsR3+MemRegsR1+RegsR2(6) direct addressing or absolute addressing: for example: ADD R1, (1001): RegsR1-RegsR1+Mem1001 meaning(7) memory indirect addressing: example: ADD R1, a (R3) meaning: RegsR1-RegsR1+MemMemRegsR3(8) the increment addressing: example: ADD R1 (R2) + meaning: RegsR1-Regs

20、R1+MemRegs(9) decrement addressing(10) zoom addressingThe function of structure design of the instruction set, 4:The instruction set classification structure in operationThe type of operation example(1) arithmetic and logical operations on integer arithmetic and logic operations: addition, subtracti

21、on, and, or etc.(2) data transmission LOAD/STORE(3) control branch, jump, procedure call and return, trap(4) operating system calls, virtual memory management.(5) floating point addition and subtraction operation(6) is converted to decimal decimal decimal decimal add, multiply, to characters(7) stri

22、ng string comparison, mobile(8) the pixel operation, compression operation5, complex instruction machine (CISC): refers to strengthen the instruction function, realize the function of software to hardware design, computer system to realize the instruction set architecture based on.The shortcomings o

23、f the CISC instruction set:(1) in the command system, the frequency of use of all kinds of orders is different.(2) the CISC instruction set architecture complexity brings complexity of computer architecture, which not only increases the development time and cost, but also easy to cause the design er

24、ror.(3) the CISC instruction set architecture complexity brings great burden to the VLSI design, is not conducive to the monolithic integration.(4) in the CISC instruction set architecture, many complex instructions require very complex operation, so slow.(5) in the CISC instruction set architecture

25、, because of the directive function is not balanced, not conducive to the use of computer architecture technology (such as advanced water technology) to improve the performance of the system.In 1980s 6, reduced instruction set computer developed: its purpose is to reduce the instruction set of the c

26、omplexity of the structure as far as possible, in order to simplify the realization of the goal of improving performance, but also in todays instruction set is a main trend of the structure and function of design.Follow the design principles:(1) choose to use the highest frequency of instruction, an

27、d added some of the most useful instructions.(2) the function of each instruction is as simple as possible, and completed in one machine cycle.(3) all have the same length as the instruction.(4) only the LOAD and STORE operating instructions to access memory and other instruction operation is perfor

28、med in the register between.(5) in a simple and effective way to support advanced language.7 operand types: integer, decimal, floating point (point), characters, strings, vectors, stack etc.There are two ways to express 8, operand types: (1) specified by the operation code encoding. (2) data can be

29、a mark by the hardware to explain the type of the operand specified by these tags, so as to choose the appropriate operation.9, the operand type size: byte (8), the word (16b), the word (32b), double word (64b)The third chapter1, pipeline technology: refers to a repeat of the timing process is decom

30、posed into several sub processes, and each process can be effective in its special function with other processes executing at the same time.2, pipeline classification: (1) according to the function of the number of points: single function pipeline, multifunctional pipeline;(2) according to the conne

31、ction between the same time segments to static and dynamic pipeline pipeline(3) according to the line level: component level pipeline (operation line), pipelined processor (instruction pipelining), inter processor pipeline (macropipeline)(4) according to whether there is water between each section o

32、f a feedback loop: linear and nonlinear pipeline pipeline(5) according to the data representation: scalar processor, vector processor3, the first processor controller structure comprises three independent controllers and four buffer stack. The three controller: memory controller, controller, controller operation instruction. Four: the first instruction buffer buffer stack stack, linear buffer stack, stack current readings, then write the number of stack.4, the throughp