or download this
C:\test>for %t in (1,2,4,16,63,255) do @interlock %t 10000000
1 threads x 10000000 iterations of normal increment non-shared to
+ok 0.269178904 seconds
...
1023 threads x 1000000 iterations of normal increment non-shared to
+ok 7.115383938 seconds
1023 threads x 1000000 iterations of interlockedIncrement16() to
+ok 127.386793294 seconds
1023 threads x 1000000 iterations of normal increment shared to
+ok 112.850957873 seconds