This is a follow on from a previous post explaining what SIMD is, how .NET developers can use it, and what performance gains they can expect to see. While the previous post showed how to take advantage of SIMD, this post will give a deeper understanding of what happens when you use it in both C## and C++ by looking at the assembly code that’s actually executed. This should help in understanding why you get the performance gains you do, why you may not get as much performance as you’d expect, and enable you to debug any problems. It also highlights the different costs and benefits of using SIMD in these languages and illustrates a few reasons why C++ should always have the upper hand in producing maximally optimised executable code.
#c++ #c #c# #simd
