Customer at Risk with No-Parity Memory? Chapter 2
Edward George Dickinson
with Gateway? Anyway, it's Gateway's contention that no-parity
memory id now the *industry standard*!
I decided to pick a few major vendors, and check. Between 1 and 1:30
Eastern time today (Wed Nov17 93) I called the following:
Dell 800-568-2902
Austin 800-752-1577
Zeos 800-554-7172
After the usual short irritant of voice-menus, I got answers from all;
they are shipping ONLY parity memory in their systems & have no plans
to change.
Fairness/honesty requires that I report completely--
CompuAdd 800-289-9888 (Greg)
This one probably doesn't rank with the other vendors; he said they
have one system board (of several?) that CAN take non-parity memory.
His answers tended to lead away from the question, I & didn't find
out how a buyer could know which he/she was getting. He also claims
that MacIntoshes have run for years with no-parity memory; maybe some-
one else could follow this up.
--Ed
David Sachs
|> He also claims that MacIntoshes have run for years with no-parity memory; maybe someone else could follow this up.
The remark about Macintoshes is correct. The ONLY Macintosh computers with parity memory have been special order varients (mostly to satisfy bid requirements) of standard models.
I do expect that eventually ECC (error code correcting) memory will become common on desktop computers. With a 64 bit wird memory bus, the simpler ECC schemes require the same number of memory bits (72) as byte parity.
Bob Nick
>Dell 800-568-2902
>Austin 800-752-1577
>Zeos 800-554-7172
>After the usual short irritant of voice-menus, I got answers from all;
>they are shipping ONLY parity memory in their systems & have no plans
>to change.
That is strange, especially since Dell is one of the companies that
started shipping non-parity even before Gateway. Perhaps your question
was general enough that they didn't really give you the correct answer.
>Fairness/honesty requires that I report completely--
>that MacIntoshes have run for years with no-parity memory; maybe some-
>one else could follow this up.
MacIntoshes have never used parity.
I'm not defending the non-parity RAM, as there is always a chance of
problems. It appears that the majority of "parity error" problems are
really false reports caused by slightly mismatched memory speeds. The
parity checking is done in hardware, so its very easy for glitches to
cause a parity error interrupt to take place while the data is actually
OK. Some programs seem even more sensitive to the false parity error
problem. There are some legitimite parity failures, and there is always
a possibility that the error could be in a data area, and go undetected.
If it is code area, the program will probably crash.
Bob
--
| Bob Nick | rn...@nyx.cs.du.edu |
| Staff Engineer | ================================================|
| SPARTA, Inc. | Disclaimer--all views are my own and do not |
| Huntsville, Al | represent the views of my employer |
Randy Marks
In article <2cdrr4$9...@snoopy.cis.ufl.edu>, e...@mole.cis.ufl.edu (Edward George Dickinson) writes...
>Was it in this newsgroup (Phil Ellis), or in my phone conversations
>with Gateway? Anyway, it's Gateway's contention that no-parity
>memory id now the *industry standard*!
>
>I decided to pick a few major vendors, and check. Between 1 and 1:30
>Eastern time today (Wed Nov17 93) I called the following:
>
>Dell 800-568-2902
>Austin 800-752-1577
>Zeos 800-554-7172
>
>After the usual short irritant of voice-menus, I got answers from all;
>they are shipping ONLY parity memory in their systems & have no plans
>to change.
I've got a fairly new Zeos 486DX-33.
I'd be glad to check it out if somebody can explain to me how I
can determine whether it has parity memory or not.
Randy Marks
(INTERNET) ma...@ssdevo.enet.dec.com
..........................................................................
"No man's life, liberty, or property is safe
while the legislature is in session."
..........................................................................