As you can see, the combination of malloc()+memset() is signifigantly
faster than calloc. (in 100M iterations of allocating a 64 byte chunk)

10:48am cmorris@server1  /home/cmorris/somedir/ #>./alloc_race
elapsed time is 43 secs (malloc&memset)
elapsed time is 51 secs (calloc)
10:50am cmorris@server1  /home/cmorris/somedir/ #>./alloc_race
elapsed time is 43 secs (malloc&memset)
elapsed time is 49 secs (calloc)
10:52am cmorris@server1  /home/cmorris/somedir/ #>./alloc_race
elapsed time is 44 secs (malloc&memset)
elapsed time is 50 secs (calloc)

10:51am cmorris@server2  /home/cmorris/somedir/ #>./alloc_race
elapsed time is 54 secs (malloc&memset)
elapsed time is 65 secs (calloc)
10:58am cmorris@server2  /home/cmorris/somedir/ #>./alloc_race
elapsed time is 55 secs (malloc&memset)
elapsed time is 65 secs (calloc)