Bandwidth vs. Throughput     Last updated on 2010     3     30, a full moon day;

After connectivity has been confirmed, protocol & port to be concerned; Analyzing topology, faster throughput result can be; RFC provides detail protocol specification, testing provides better engineering;

Bandwidth and throughput have been used for several decades already; WHEN 2 computing-nodes are being connected and the 2 computing-nodes are in a communication, WHERE FC, Fast Ethernet, and etc. define WHAT kind of communication's medium, and WHICH server serves higher performance than others;

Within 1 bandwidth, in common, several frequencies exist, those several frequencies have been modulated, and mux [aka multiplexed] within the bandwidth; Also see: FDM, TDM, ... ; WHEN several bandwidths are needed, and then scalability begins; Scale/Scalable/Scalability in computing refers HOW to manage several bandwidths; In Grid vs. Others, knowledge of system analysts and engineers make differences among existing systems' performance based on how to manage the systems' scalability and workloads;

Bandwidth and throughput are syntax vs. semantic alike;

For example, 1 bandwidth can be 10 Gbps [Mostly by marketing ad], but after attenuation, CRC, Resistance, S/N ratio, ... , the bandwidth might perform between 2 computing nodes' communication about 1Gbps; Generally, bandwidth is a speed of communicating between 2 nodes, but in common, perform much lesser than ... ;


Throughput defines scientific performance and accomplishment; For example, 1 throughput can be 8Gbps [i.e. 8Gbps per 80TB single FS ], including bandwidths' disturbances [i.e. attenuation, dB/m, CRC, Lag time, Ohn cm, Propagational delay time, S/N ratio, ...]; Therefore, usage "bandwidth" does not define performance, on the other hand, throughput does;