The performance rate of a computer is how much useful work it can accomplish compared to how long it runs and the resources used. More work with less time and resources makes for a better-performing machine. This is quite intuitive. More specifically, there are several metrics of good performance in computer science, as follows:
- fast data compression/decompression
- quick response times
- high throughput
- high availability of computing system/application
- low resource usage
- high bandwidth/short data transmission time
The performance of any computer system can be evaluated in measurable technical terms using one or more of the metrics. This was the performance can be compared relative to other systems or the same system before/after changes. It can also by defined in absolute terms.
IBM's Blue Gene project creates some of the fastest supercomputers in the world. [Photo credit: Argonne National Laboratory]
There are a number of technical performance metrics that indirectly affect overall computer performance. The total amount of time (t) required to execute a particular benchmark program is
t = N * C/f
or
P = I * f/N
Where:
- P = 1/t is the performance in terms of time-to-execute
- N marks the number of computer instructions which are executed
- f stands for the clock frequency, in clock cycles/second
- C = 1/I is the average cycles per instruction