Showing posts with label Floating-point. Show all posts
Showing posts with label Floating-point. Show all posts

Pain Killers for the Fixed-Point Design Flow


Taking a floating-point representation of an algorithm into a fixed-point representation is an integral step on the path towards implementation. Unfortunately, it is known to be time-consuming, tedious and error prone. This paper from synopsys describes proven methodologies on how to overcome the hurdles, including many useful tips to get you to results faster.