LINFO

Throughput Definition



Throughput is the amount of work that can be performed or the amount of output that be produced by a system or component in a given period of time. It has a meaning similar to that of capacity, and the two are often used as synonyms.

The term is used in a variety of contexts. For example, in a transportation system it can refer to the amount of freight or passengers that can be carried by a railroad line or the amount of a liquid or gas (e.g., petroleum or natural gas) passing through a pipeline during some specified time period. In the case of manufacturing, throughput can refer to the amount of steel that could be produced by a steel mill or to the amount of an antibiotic that could be produced by a pharmaceutical plant, both per day, month, year, etc.

In a computer context, throughput is the amount of work that a computer can do in a given period of time. The work can be measured in terms of the amount of data processed or transferred from one location to another by a computer, computer network or computer component. Data transfer rates for disk drives and networks are commonly measured in terms of kilobits per second (kbps), megabits per second (Mbps) or gigabits per second (Gbps).

The throughput of a computer depends on such factors as the speed of the CPU (central processing unit), the amount of memory, the performance of the operating system and the means of measurement. The throughput for a network additionally depends on such factors as the type of transmission media (e.g., twisted pair copper wire cable, coaxial cable, optical fiber cable or radio waves), the speed of the routers and other network devices, and the network protocol.

The term scalable refers to the situation in which the throughput increases roughly in proportion to the increase in the inputs. For example, if blast furnaces can be manufactured in a variety of sizes and if a larger sized blast furnace will produce proportionately more steel of the same quality and at the same cost (and a smaller blast furnace will produce proportionately less steel of the same quality and cost), then the blast furnace technology is scalable. Likewise, a computer operating system is scalable if it can perform just as efficiently on a supercomputer or on a large network of computers as it can on a single, stand-alone desktop computer.






Created March 2, 2006.
Copyright © 2006 The Linux Information Project. All Rights Reserved.