Slide 3.1: Defining performance

Slide 2.23: Chip manufacturing process (cont.)
Slide 3.2: Defining performance (cont.)
Home Print version

Defining Performance

Computer performance is determined by different ways and is usually related to the following two terms:

Response/execution time, which is the total time required for the computer to complete a task, including disk accesses, memory accesses, I/O activities, operating system overhead, CPU execution time, etc.

Throughput/bandwidth, which is another measure of performance. It is the number of tasks completed per unit time.

Question: Which one is improved if the processor in a computer is replaced with a faster version?

Answer: response time throughput

Question: Which one is improved if additional processors are added to a system that uses multiple processors for separate tasks?

Answer: response time throughput

The textbook is primarily concerned with response time. To maximize performance, try to minimize response time or execution time for some tasks. Thus, performance is related to execution time for a computer X:

    Performance_X = 1 / Execution time_X

This means that for two computers X and Y, if the performance of X is greater than the performance of Y, we have

    Performance_X > Performance_Y

 ⇒ ( 1 / Execution time_X ) > ( 1 / Execution time_Y )

 ⇒ Execution time_Y > Execution time_X

i.e., the execution time on Y is longer than that on X, if X is faster than Y.

◀
Previous

Slide 2.23: Chip manufacturing process (cont.)
Slide 3.2: Defining performance (cont.)
Home Print version

▶
Next

Wife: “I want another baby.”
Husband: “That’s a relief, I also really don’t like this one.”