Measuring and reporting performance

Ø  When we say one computer is faster than other? ·         User say a computer is faster when program runs in less time. ·         Administrator tells a computer is faster when it completes more transaction per hour. ·         RESPONSE TIME: The time between the start and the completion of an event is known as response time. ·         THROUGHPUT: Total amount of work done in a given time. EXAMPLE    Consider two computers X and Y.Then ‘X’ is faster than ‘Y’ when Execution time is lower on X than on Y. ·         X is ‘n’ times faster than Y n         = Execution time of y              ---------------------------              Execution time of x        Execution time is the reciprocal of performance                          n        =   Performance x                                       -----------------------                                         Performance y             WHAT IS CPU TIME Ø  When processos is computing one program, waiting time of I/O is present which minimize response time . so cpu time is recognized which doesn’t include waiting time for I/O.             v       BENCHMARKS Ø  Best choice of benchmark to measure performance.attempts at running program that are much simpler than real application have led to performance pitfalls. Ø  TYPES ·         Kernel: small,key piece of real application. ·         Toy program:its a 100 line program from beginning programming    assignment,ex:quick sort. ·         Synthetic benchmarks:take program invented to try to match the behaviour of real application. v  There Are Three Different Approaches To Addressing Whether Source Code Modifications Are Allowed: 1. No source code modifications are allowed. 2. Source code modifications are allowed, but are essentially impossible. For example, database benchmarks rely on standard database programs that are tens of millions of lines of code. The database companies are highly unlikely to make changes to enhance the performance for one particular computer. 3. Source modifications are allowed, as long as the modified version produces the same output. v  Benchmark suites Ø  The source code modification shouldn’t affect the accuracy, so Benchmark suite is used. Ø  It’s a popular measure of performance of processor with variety of application. v  Desktop benchmarks Ø  Processor Intensive Benchmark and Graphics Intensive Benchmark are present here. Ø  SPEC: create benchmark focusing on processor performance. Ø  In SPEC real program modified to be portable & minimize effect of I/O on performance.    v  Server benchmark Ø  Have many functions. Ø  Simplest in processor throughout oriented benchmark. Ø  TRANSACTION PROCESSING measure ability of system to handle transations. Ø  SPEC rate can be computed from                               Geometric Mean = n v-n i=1 samplei