Re: Nine bits v. Eight



 > Date: Tue, 25 Apr 1995 17:08:57 -0400 (EDT)
 > From: Friedman <pfriedma()at()acnet.pratt.edu>
 > Subject: CCL:Nine bits v. Eight
 >
 > We are purchasing a pentium 100MHZ PC to speed up our runs using Hyperchem
 > and Gaussian 92W/DFT. I was alerted by several people that many of the
 newer
 > PCs use 8 bit SIMMs now. I've gotten differing opinions on whether parity
 > checking is needed. I would appreciate hearing from anyone with an informed
 > opinion as to the risks (no pun) involved. Hypercube and Gaussian boffins
 > especially welcome.
 I relayed the question to our hardware expert, Edmund Lam.  He responded:
 ---------
 In the past, PC used 9 bit wide RAM.  This architecture was used to ensure
 detection of RAM errors.  In recent times, the quality of RAM and mainboards
 have curved the need for parity detection.
 MACs have always run 8 bit wide RAM.  Workstations such as the Sparcs also
 use 8 bit wide RAM.
 The extra bit can be used to detect a memory error, but is not sufficient to
 correct it.
 ----------
 My own experience with parity checking was a PC that I ended up lugging
 back and forth to and from the shop for nearly a year because it kept
 giving me "Parity error -- system halted" crashes at home but worked
 fine at
 the shop.  Shop kept claiming it was a problem with the RAM, and replacing
 the chips.  Took them a year to clue in that the problem was actually
 with the motherboard; after they replaced *that*, the machine gave no
 more trouble.  Whether that little story argues FOR or AGAINST the merits
 of parity-checking is something I'll leave for others to sort out.  :-)
 Regards,
 Joel
 ------------
 Joel Polowin, Ph.D.   Manager, Scientific Support
 Email to: polowin()at()hyper.com
 Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040
 Info requests to: info()at()hyper.com    Support questions to:
 support()at()hyper.com
 Email group: Send "subscribe hyperchem" to
 hyperchem-request()at()hyper.com
 WWW: http://www.hyper.com/