The Intel JIT should implement the following optimizations:
  1) Small final methods should be inlined.
  2) Loops should be aligned on 16-byte boundaries.
  3) Integer multiply by a power of two should be done as a shift.