My question is about the difference between Vector and superscalar processors?
Of course they differ in programming but are we gonna see them on PCs?
Of course they differ in programming but are we gonna see them on PCs?
A question for Engineers on Lebgeeks.com
4 days later
The Scalar Processors are the ones found in PC's today. No ?
Yup i know these stuff the ps3 has 7 SPE or vector processors and one SuperScalar but i meant the difference...it seems no one here knows.
Well i don't know much about them but your can read here
http://en.wikipedia.org/wiki/Vector_processor
http://en.wikipedia.org/wiki/Scalar_processor
http://en.wikipedia.org/wiki/Vector_processor
http://en.wikipedia.org/wiki/Scalar_processor
The difference is easy to understand , ill give you an example .
assuming you don't know anything about programming ; i have a program that needs to both multiply a*b and divide C/D .
in a scalar(most popular processor) works by the FIFO principle "first in first out" . so naturally its going to multiply a*b and dump them the result in a separate cell , then divide C/D and dump the result in another separate cell .
while in vector processors , its able to process two or more simultaneous orders and processes them at the same time . it can do that as long as one condition is satisfied , the variable being worked with needs to be available and not in use , in order to not create data loss or false data . so A/B A*B will be worked with on a scalar processor the same thing they are going to be dealt with on a vector . another added advantage is that in vector processing there is no time lost to read , fetch , decode data . it does it once then react according to all operands . while in a scalar it reads , fetches , then decodes for every single operand .
any questions class ? :P
assuming you don't know anything about programming ; i have a program that needs to both multiply a*b and divide C/D .
in a scalar(most popular processor) works by the FIFO principle "first in first out" . so naturally its going to multiply a*b and dump them the result in a separate cell , then divide C/D and dump the result in another separate cell .
while in vector processors , its able to process two or more simultaneous orders and processes them at the same time . it can do that as long as one condition is satisfied , the variable being worked with needs to be available and not in use , in order to not create data loss or false data . so A/B A*B will be worked with on a scalar processor the same thing they are going to be dealt with on a vector . another added advantage is that in vector processing there is no time lost to read , fetch , decode data . it does it once then react according to all operands . while in a scalar it reads , fetches , then decodes for every single operand .
any questions class ? :P
thnx Flakk i already read them but they are for programmers.Well i don't know much about them but your can read here
http://en.wikipedia.org/wiki/Vector_processor
http://en.wikipedia.org/wiki/Scalar_processor
So Jad explained it meaning that Vector pressecors are best suited for Heavy loads like gfx and Superscalar for for general purpose.thnx Jad...
- Edited
Thanks for the explanation.
But dont think Jad ever worked in Vector processors a lot. PS3 has the SPEs which are Vector processors.