第一节系统结构概述.pptVIP

Computer Architecture 3.0 Computer Architecture 3.0 Example1.4: Suppose we have made the following measurements: Frequency of FP operations (other than FPSQR) = 25% Average CPI of FP operations = 4.0 Average CPI of other instructions = 1.33 Frequency of FPSQR= 2% CPI of FPSQR = 20 Assume that the two design alternatives are to decrease the CPI of FPSQR to 2 or to decrease the average CPI of all FP operations to 2.5. Compare these two design alternatives using the CPU performance equation. Computer Architecture 3.0 Computer Architecture 3.0 Principle of Locality(局部性原理) Programs tend to reuse data and instructions they have used recently. a program spends 90% of its execution time in only 10% of the code. （一个程序90％的时间都在执行它10%的代码） Temporal locality （时间局部性） states that recently accessed items are likely to be accessed in the near future. Spatial locality (空间局部性) says that items whose addresses are near one another tend to be referenced(访问) close together in time. Computer Architecture 3.0 MIPS (million instruction per second) 3 MIPS的问题: MIPS 与指令集相关, 这个使得MIPS架构计算机和其他不同指令集计算机比较困难度增加 在同一计算机上 对于不同的程序MIPS是不同的 最重要的是，, MIPS 与性能可能是相反的! The bigger the MIPS, the faster the machine. Computer Architecture 3.0 Fallacies and Pitfalls Fallacy MIPS is an accurate measure for comparing performance among computers. MIPS是比较机器间性能的准确指标 The bigger the MIPS, the faster the machine. Classic example: MIPS rating on a machine with floating point hardware may be LOWER than machine with Software floating point because Software floating point executes simpler instructions ,resulting in a higher MIPS rating. MIPS is typically used to measure PEAK performance(最佳性能), not real performance. Computer Architecture 3.0 Classic example: The figure also shows the performance of a 1.7 GHz Pentium 4 relative to a 1 GHz Pentium 3. In all cases except the SPEC floating-point suite, the Pentium 4 delivers less performance per MHz than the Pentium III. 谬论和缺陷 谬论 同一指令集系统(ISA)的2个处理器的性