If I have a long string of commands in a pipeline, will getting the result always be faster than if using the same commands divided up? How is it faster or slower? What is the overhead for pipelines versus splitting the commands up into steps.

Also, does using multicore processors give you a large benefit from a long pipeline?


I am thinking of something like: cat file | gzip it up OR cat file | gzip decompress >> something OR humor me with cat file | gzip it up | gzip decompress | gzip it up | gzip decompress

I am very interested in this. I have taken a computer architecture class but am just really getting into linux. Thanks.