This section investigates the effects of the different code optimizations we implemented for the FastVM on x86. These optimizations are register allocation, method inlining, and floating-point precision mode optimizations as we described in section 3. All three optimization techniques impose a negligible amount of additional compilation time to the VM.