From: mayan@bestweb.net   
      
   >>Maybe you're looking for papers on cycle-accurate simulation of   
   >>processors.  If those are not useful, at least they will provide you   
   >>with pointers to (hopefully) what you really want.   
   >   
   >   
   > However, you should be aware that no simulator exists till now that has   
   > been validated by statistics. The simulation errors are unfortunately   
   > not neglectible.   
      
   Not exactly true: see "Validation of Turandot, a fast processor model   
   for microarchitecture exploration", IPCCC 1999.   
      
   The Turandot simulator was within 5% for the Power-4   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)