XPost: sci.space.policy, sci.physics, sci.electronics.design   
      
   In sci.physics Jeff Findley  wrote:   
   > In article , krw@notreal.com   
   > says...   
   >> The point is that CAD on minicomputers was very minimal.  It was the   
   >> domain of the mainframe.  Rubylith was the tool of choice for the   
   >> electronics industry.   
   >>   
   >   
   > I've worked on CAE software that's tightly integrated with CAD my entire   
   > professional life.  Back in about 1988 our CAD/CAE software still ran on   
   > mainframes (IBM, DEC, and etc.) but the transition to Unix workstations   
   > was in its infancy.  Back then, PCs were "toys" that quite simply   
   > couldn't handle professional level CAD/CAE software.   
   >   
   > In the early 1990s Unix Workstations dominated for running CAD/CAE   
   > software.  A good SGI "box" would run you about $20k in early 1990s   
   > dollars (about $33k today).   
   >   
   > Today, you can comfortably run CAD/CAE software (at least the CAE   
   > pre/post) on a sub $2k PC running Windows OS.  But many customers will   
   > go quite a bit over $2k with things like solid state drives and 64 GB or   
   > more of RAM coupled with the best professional graphics card money can   
   > buy (no, they're not quite the same as consumer/gaming cards).  Still,   
   > the most "decked out" PC workstation today will still cost a fraction of   
   > what a Unix workstation used to cost in the early 1990s.   
   >   
   > So again, we see yet another example of improving technologies driving   
   > down costs in a market.   
      
   Nope, what we see is yet another example of consumer demand driving down   
   manufacturing costs by encouraging high volume, automated, manufacturing.   
      
   --   
   Jim Pennino   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)