John Walker's FBENCH: Mono vs .NET

I implemented John Walker's FBENCH today in C# to see how Mono did compared to MS.NET. I had always thought that Mono's integer arithmetic was faster than MS.NET, but these test results say differently.

This is the benchmark being run in .NET. The times were about as consistent as you can get.

This is Mono running the benchmark (again, times were extremely consistent).

The benchmark did 1000 iterations of John Walker's raytracing FBENCH. Source code can be found here. This is my first try at something like this, so if you see something I did wrong, please let me know. Thanks.