From: anton@mips.complang.tuwien.ac.at   
      
   John Levine  writes:   
   >Apropos another thread I can believe that IA-64 was obsolete before it was   
   shipped   
   >for that reason, static scheduling will never keep up with dynamic except in   
   >applications where the access patterns are predictable.   
      
   Concerning the scheduling, hardware scheduling looked pretty dumb at   
   the time (always schedule the oldest ready instruction(s)), and   
   compilers could pick the instructions on the critical path in their   
   scheduling, but given the scheduling barriers in compilers (e.g.,   
   calls and returns), and the window sizes in current hardware, even   
   dumb is superior to smart.   
      
   Another aspect where hardware is far superior is branch prediction.   
      
   >Are there enough applications like that to make VLIWs worth it?  Some kinds   
   of DSP?   
      
   There have certainly been DSPs from TI (C60 series IIRC), and Phillips   
   (TriMedia) that have VLIW architectures, so at least for a while, VLIW   
   was competetive there.   
      
   - anton   
   --   
   'Anyone trying for "industrial quality" ISA should avoid undefined behavior.'   
     Mitch Alsup,    
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)