home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.arch
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!cs.utexas.edu!torn!utzoo!henry
- From: henry@zoo.toronto.edu (Henry Spencer)
- Subject: Re: BUSES
- Message-ID: <BrusoI.2so@zoo.toronto.edu>
- Date: Thu, 23 Jul 1992 18:00:16 GMT
- References: <1992Jul23.092211.18462@nuscc.nus.sg>
- Organization: U of Toronto Zoology
- Lines: 21
-
- In article <1992Jul23.092211.18462@nuscc.nus.sg> eletanjm@nuscc.nus.sg (TAN JIN MENG) writes:
- >It seems to me that many if not most applications do not need (relatively
- >speaking) such a high bandwidth...
- >Software controls the rate at which you exercise the bus at these high speeds.
-
- True. However, you assume that software efficiency is remaining constant.
- It's not; it's declining. Even without considering the rising expectations
- of the customers, software bloat is more than keeping pace with the hardware.
- Running X at useful speed takes far more horsepower than, say, the old Blit
- window system (typically much snappier than X despite running on a 68000).
- If you're one of the poor people doomed by bureaucratic fascism to run the
- OSI networking protocols, you need far more horsepower than the TCP/IP crowd
- to get the same performance. And so on.
-
- >Also, if I get my file within 0.5 sec as opposed to 1.0 sec, will I
- >notice enough to pay for the hardware?
-
- If you get a hundred files in 50 seconds as opposed to 100, you will notice.
- --
- There is nothing wrong with making | Henry Spencer @ U of Toronto Zoology
- mistakes, but... make *new* ones. -D.Sim| henry@zoo.toronto.edu utzoo!henry
-