XPost: alt.comp.hardware.overclocking   
   From: hgoldste@mpcs.com   
      
   On Thu, 11 Dec 2008 08:26:37 -0500, Phil Weldon    
   wrote:   
    :  'Howard' wrote:   
    : > Can you fillin a blank please? What's the application?   
    :  _____   
    :   
    :  Gedanenexperiment   
      
   I was sort of hoping there'd be a use for it   
      
    :   
    :  Phil Weldon   
    :   
      
      
      
    :  "Howard"  wrote in message   
    :  news:slrngk24qe.mnv.bit-bucket@individual.net...   
    : > On Wed, 10 Dec 2008 10:21:41 -0800 (PST), omattos    
    : > wrote:   
    : > : > This kind of redundancy is always be more expensive in time and   
    : > material   
    : > : > than reducing the error rate by merely operating components 'in spec'.   
    : > :   
    : > :  I was more thinking of redundancy like this for performance.  Say I   
    : > :  have a task that must be completed in 1 second, but takes 2 seconds on   
    : > :  the best currently available processor.  The task can't be   
    : > :  parallelized (ie. every part of the task depends on the previous).  By   
    : > :  using multiple processors running faster than designed to get the task   
    : > :  done I can get the job done in 1 second.  By having multiple   
    : > :  processors running the same job simultaneously, I can check which   
    : > :  result is correct (by a majority vote)   
    : > :   
    : > :  This uses the theory that as the clock speed of the part moves out of   
    : > :  it's designed working region the reliability goes down, and the   
    : > :  further out of the working region, the further it goes down.   
    : > :   
    : >   
    : > Can you fillin a blank please? What's the application?   
    :   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)