Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!news.uu.net!unhd.unh.edu!IceCube.unh.edu!tjb
From: tjb@IceCube.unh.edu (Thomas J. Baker)
Newsgroups: comp.sys.next
Subject: Performance Penality of Parity Memory?
Message-ID: <1991May22.111900.5785@IceCube.unh.edu>
Date: 22 May 91 11:19:00 GMT
Sender: tjb@IceCube.unh.edu (Thomas J. Baker)
Reply-To: tjb@IceCube.unh.edu
Organization: tjb consulting, dover NH
Lines: 26

NeXTanswers states under caveats of parity memory:

"There is a speed penalty for accessing parity memory, since for each access a  
parity check sum must be computed.  This translates into slower overall  
performance."

Can anyone speak from experience about this performance degradation?  It's time  
to buy some 4Mb simms and I figured someone out there must by now have some  
parity memory installed.  It is noticable?  Is it 10%?  

Thanks for any info...

tjb

--
+-------------------------------------------------------------------+
| Thomas Baker                           Voice: (603) 743-4987      |
| P.O. Box 213                                                      |
| Durham, NH 03824           Internet/NeXTmail: tjb@IceCube.unh.edu | 
+-------------------------------------------------------------------+
-- 
+-------------------------------------------------------------------+
| Thomas Baker                           Voice: (603) 743-4987      |
| P.O. Box 213                                                      |
| Durham, NH 03824           Internet/NeXTmail: tjb@IceCube.unh.edu | 
+-------------------------------------------------------------------+