In a nutshell, Cp is a measure of your process variation compared to your spec limits and Cpk is measure of both variation and how well your process is centered within your spec limits.
Cp = (USL-LSL)/6*sigma
where USL=upper spec limit LSL=lower spec limit and sigma=standard deviation
Cpk is the smaller of two calculations: Cpklower = (mean-LSL)/3*sigma Cpkupper = (USL-mean)/3*sigma
We don't use Cp much these days. We used to use it a lot back in the early days of placement equipment. For example, if a piece of equipment had a good Cp we were happy because we knew we could tweak the mean to center the distribution. These days we expect equipment to both have low standard deviation and be centered within the spec limits so we tend to ignore Cp and use Cpk.
Here's a good visual: http://elsmar.com/Cp_vs_Cpk.html
reply »