Efficiency of vector processing over scalar processing, Computer Engineering

Assignment Help:

Efficiency of Vector Processing over Scalar Processing

As we know a serial computer processes single scalar operands at a time. So if we have to process a vector of length 'n' by the sequential computer then vector should be divided into n scalar steps and implemented one by one.

E.g. think about the following vector addition:

A + B → C

The vectors are of length 500. This operation by the serial computer can be specified by 500 add instructions as written below:

969_Efficiency of Vector Processing over Scalar Processing.png

If we execute the same operation by a pipelined vector computer then it doesn't break the vectors in 500 add statements. Since a vector processor has the set of vector instructions which allow the operations to be specified in one vector instruction like:

A (1:500) + B (1:500) → C (1:500)

Each vector operation can be broken internally in scalar operations however they are executed concurrently which yields in much faster execution as compared to serial computer. So the benefit of using vector processing over scalar processing is it eliminates the overhead caused by the loop control needed in a serial computer.

Related Discussions:- Efficiency of vector processing over scalar processing

Write Your Message!

Name

Email id

Message

Verfication Code

User Account

All Pages