Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!olivea!orc!bu.edu!snorkelwacker.mit.edu!ai-lab!wheat-chex!caroma
From: caroma@wheat-chex.ai.mit.edu (Carl R. Manning)
Newsgroups: comp.sys.next
Subject: Re: memory (was: slab vs. cube and $$)
Summary: Parity isn't the same as ECC; what happens when a parity error occurs?
Message-ID: <12108@life.ai.mit.edu>
Date: 30 Nov 90 16:27:57 GMT
References: <12081@life.ai.mit.edu> <1990Nov29.161041.15775@magnus.ircc.ohio-state.edu> <1990Nov30.013031.25032@mp.cs.niu.edu>
Sender: news@ai.mit.edu
Lines: 45

In article <1990Nov30.013031.25032@mp.cs.niu.edu> bennett@mp.cs.niu.edu (Scott Bennett) writes:
>In article <1990Nov29.161041.15775@magnus.ircc.ohio-state.edu> tgingric@magnus.ircc.ohio-state.edu (Tyler S Gingrich) writes:
>>In article <12081@life.ai.mit.edu> caroma@rice-chex.ai.mit.edu (Carl R. Manning) writes:
>>>Memory is cheap; add a few more bits and you can have ECC error
>>>correction of single bit errors.  Perhaps futures NeXTs can be
>>>designed for people or servers who need this much reliability, at
>>>least as an option.
>>
>>You can get ANY of the Next computers with Parity memory.  
>
>>someone else writes:
>>>Well actually there are several nexts available with parity memory.
>>>its an option, if you want.  Of course, it costs extra
>>>All four machines are available with parity.  
>>

Some people seem to have jumped onto this thread late, so I'll repeat
and reword parts of my posting:

I realize that the NeXT can have parity memory; my original question
remains unanswered: What does the NeXT do when it gets a memory parity
error?  (e.g. Does it lock up the machine immediately?  Or does it report
the error and allow you to decide whether to try continuing?  If the
error is in memory which isn't critical to what I'm doing, I'd rather
not lose all my work with a crashed machine -- at least give me the
chance to try to save what work I can.)

I have a feeling some people may be missing part of the point of my
posting.  As the amount of memory increases, the chances go up that a
bit will get flipped by a passing cosmic ray or that some chip or part
of a chip will go bad.  Parity memory adds one bit to each word, which
allows detection of single bit errors.  ECC (error correction code)
memory adds a few more bits, which add enough redundancy so that words
can be correctly reconstructed even if some bit(s) go bad.  A common
form of ECC allows single bit error correction and double bit error
detection.  Thus, it is possible to organize memory so that a passing
cosmic ray or single failing chip needn't stop your machine; this
security can be an important selling point for people who depend on
their machine to accomplish their work.

So I suggest the next generation NeXT's might offer the option of
error correcting code memory rather than just parity.  Ordinary PC's
offer parity; more `premium' PC's and workstations offer ECC.

-- CarlManning