General term of throughput is how much work is completed or how much output is produced over specific period of time. However, the interpretation of the term may differ in a various contexts, such as manufacturing, computing, and networking. "Digital bandwidth consumption" can be interchangeably used as throughput.

 

In a manufacture, throughput is how much output is produced in a unit period under normal operating condition. Examples are the output of steel that a steel mill could produce or the output of an antibiotic that a pharmaceutical plant could produced per day, month, year, etc.

 

In a computer, throughput is a measure of system’s performance in a specific period of time to send data through computer components, such as processor, computer network, and storage devices. Kilobits per second (kbps), megabits per second (Mbps) or gigabits per second (Gbps) are commonly used as a measure of data transfer rates for hard disk drives. In addition, the speed of the CPU (central processing unit), the amount of memory, and the performance of the operating system affect the throughput of a computer, and transmission media, the speed of the routers and other network devices, and the network protocol affect the throughput for a network

 

In data communications, throughput is the data transfer rate of information over network. The speed of the rate will be influenced by bandwidth, line congestion, and error correction..

 

Maximum throughput is an ability that can achieve the highest performance. “Digital bandwidth capacity” is often used as synonymous of maximum throughput. Today, maximum throughput is major interests for users of telecommunication devices, systems designers, and researchers, since it is a great measure of efficiency and effectiveness. There are four different values in terms of maximum throughput on the purpose of comparing “the upper limit conceptual performance of multiple systems”: 1) “maximum theoretical throughput” 2) “maximum achievable throughput” 3) “peak measured throughput” 4) “maximum sustained throughput.”

 

“Maximum theoretical throughput” is the largest attainable quantity of data that can be sent under ideal conditions, and its number has close connection to the channel capacity of the system. The primarily reason for the use of the maximum theoretical throughput is a rough calculated value that is used for determination of bounds on possible performance early in a system design stage.

 

“Maximum achievable throughput” of number is closely related to maximum theoretical throughput, but they are different from each other: maximum theoretical throughput uses only data packets on a channel whereas maximum achievable throughput uses both handshake and control packets that decrease the amount of channel space available for data packets and data packet length. Maximum achievable throughput is considered as a bright point of view of network performance, but it is more useful tool for understanding expected system performance than maximum theoretical throughput and is widely used in an active area of research.

 

“Peak measurement throughput” is measured in a real, implemented system, or a simulated system, and the value of throughput is measured over short period of time. Instantaneous throughput is interchangeably used as peak measured throughput. Peak measurement is a good tool of measurement for systems which rely on burst data transmission but isn’t a good tool of measurement for system performance.

 

“Maximum sustained throughput” is related to peak measurement throughput in terms of contradiction. Unlike peak measurement throughput, maximum sustained throughput of value is measured over a long period of time and a great tool of measurement for system performance because of “high duty cycle networks”.

 

Factors affecting the data throughput rate

1)    Hardware: The disk drive, the tape drive, the disk controller, the SCSI, or the improper cabling/termination are the factors of the speed of the disk controller and hard ware errors and lower performance so that proper set up and the check of the controller for the tape backup hardware are essential.

2)    System: Fragmented disks should be avoided, since it’s not only harm the rate at which data is written to tape, but also harm the overall performance.

3)    Available memory: insufficient memory, improper page file settings, and a lack of available free hard disk space are also factors of slow performance.

4)    Size and number of files: for faster backups, the disk should contain a few large size files, since it will be helpful for backups more efficiently than backing them up from multiple locations.

http://en.wikipedia.org/wiki/Throughput