SUMMARY: IPX and ELC SIMM compatibility

From: Michael Maciolek (mikem@juliet.ll.mit.edu)
Date: Thu Jan 16 1992 - 22:57:43 CST


Original quesitons:
+--------------------------------------------------------------
| 1. Can you really put 4Mx9 SIMMs into an ELC or IPX?
| 1a. Do you have to put in 4, or can you put in just 1 or 2?
|
|
| 2. Are the other SIMMs (501-1698 and 501-1822) the 1Mx33 and 4Mx33 type?
| 2a. Can you combine 1Mx33 and 4Mx33 SIMMs in the same ELC or IPX?
|
| 3. What kind of SIMMs are in a base-configured ELC (which has 8 Meg)
| or IPX (which has 16 Meg)?
|
| 4. What is the source of your information? First-hand experience ("I've
| tried it myself, and it works."), technical manual, hearsay or rumor?
+--------------------------------------------------------------

If you fanatically save all postings to this group, or if you have a good
enough memory, you may recall that my original posting included a memory
compatibility table that I copied from a piece of paper I got from my Sun
sales rep.

Well, as many of you pointed out, there was a major typo in the table; I
had interchanged some of the IPC data with SLC data. Although the typo was
not relevant to my questions (which were specifically about the IPX and ELX)
it led to some confusion, so I have enclosed at the end of this summary a
corrected table, which includes corrections and other information I received
from a number of sun-managers readers.

Now, to the answers:

1. a resounding NO, you CAN'T put 4Mx9 SIMMs into an ELC or IPX, for the
simple reason that they don't fit. The sockets are a different size. Oddly
enough, one person claimed that he had put 4Mx9 SIMMs into his IPX, but so
many others claimed that this was physically impossible because the sockets
were different sizes, I believe that this one person was talking about 1Mx33
SIMMs and got confused.

2. There was some uncertainty as to whether these parts were 33 bits wide
or 36 bits wide (perhaps they really have 36 bits, but the Sun MMU only pays
attention to 33 of them?) Anyhow, the majority tended to favor 33 bits, and
one response from Steve Campos - a technician at Highland Digital (where they
actually SELL these components) said 33 bits, so I'll say 33 bits for the
remainder of this summary, even though I got mixed signals from this list.
Steve also said that while some SLCs shipped with 100ns parts, that the
newer/faster ELC and IPX needed 80ns. This is noted in the table below.

2a. Mixing & Matching 1Mx33 with 4Mx33 SIMMs is OK. poffen@sj.ate.slb.com
(Russ Poffenberger) pointed out that the 4Mx36s should be installed in the
first slot(s), then the 1Mx33s in the remaining slot(s).

3. The ELC comes with two 1Mx33 SIMMs installed, for a base configuration
of 8 meg. Early IPXs shipped with all four slots filled with 1Mx33 SIMMs
to give 16 megabytes, but all current and future IPXs are shipping with a
single 4Mx33 SIMM, so they have 16 Meg with 3 free slots.

Here's the updated/corrected SIMM compatibility table, which reflects the
majority-response to my query:

Note - one respondent added some information for the 600MP series, and he
indicated that the 600 would accept either 80ns or 100ns 4Mx9 SIMMs, and
was jumper-selectable. This is encouraging, as the "jumper-selectable" part
implies that you can tell the board to run slower or faster - so you might
actually gain some speed advantage by using 80 ns parts if available, or you
can go the cheap and dirty route and use slower 100ns parts at some speed
penalty.

I checked this information against the technical information I have about
the 600MP series, and I've appended two lines for the 600MP and its 164A
expansion board to the bottom of this chart.

This chart is now in agreement with the majority of the responses I received,
and I believe it's correct, but standard disclaimers apply.

Thanks for the help!

                                          SIMM type & part number
----------------------------+---------+-----------------------------+---------+
                            | 1 MB | 4 MB | 16MB |
---+-------+--------+-------+---------+---------+---------+---------+---------+
S t| | Memory | | 100ns | 80ns | 100ns | 100ns | 80ns ? |
y y| Total | in base| Max. | 1Mx9 | 4Mx9 | 4Mx9 | 1Mx33 | 4Mx33 |
s p| memory| config | Memory| | | 501-1625| | |
  e| slots | (MB) | (MB) | 501-1679| 501-1739| 501-1812| 501-1698| 501-1822|
---+-------+--------+-------+---------+---------+---------+---------+---------+
IPX| 4 | 16 | 64 | | | | * 80ns | X |
---+-------+--------+-------+---------+---------+---------+---------+---------+
ELC| 4 | 8 | 64 | | | | * 80ns | X |
---+-------+--------+-------+---------+---------+---------+---------+---------+
SLC| 4 | 8 | 16 | | | | X | |
---+-------+--------+-------+---------+---------+---------+---------+---------+
SS2| 16 | 32 | 128 | | X | | | |
---+-------+--------+-------+---------+---------+---------+---------+---------+
IPC| 12 | 8 | 48 | X | X | X | | |
---+-------+--------+-------+---------+---------+---------+---------+---------+
S1+| 16 | 8 | 64 | X | X | X | | |
---+-------+--------+-------+---------+---------+---------+---------+---------+
600| 64 | 64 | 128 | | X | X | | |
---+-------+--------+-------+---------+---------+---------+---------+---------+
164| 128 | 64 | 256 | | X | X | | |
---+-------+--------+-------+---------+---------+---------+---------+---------+

notes: "S1+" = "SS1+"
        "600" = "SS 600 MP",
        "164" = "164A ECC Memory expansion board"

* The ELC and IPX systems require an 80ns 1Mx33 SIMM, not the 100ns SIMM
   listed in the column heading. No part number was available.

Responses were received from:

        Brent Alan Wiese <brent@crick.ssctr.bcm.tmc.edu>
        Bernard Silver <bs30%sirius@gte.com>
        eckhard@ikarus.ts.go.dlr.de (Eckhard Rueggeberg)
        kjl@ray.fi (Kari J Leino)
        john@rod.mitre.org (John Marsh)
        fred@johnd.tamu.edu (Fred Seals 5-2298)
        Steve Hanson <hanson@pogo.fnal.gov>
        matt@wbst845e.xerox.com (Matt Goheen)
        Richard Elling <Richard.Elling@eng.auburn.edu>
        feldt@phyast.nhn.uoknor.edu (Andy Feldt)
        jallen@nersc.gov (John Allen)
        poffen@sj.ate.slb.com (Russ Poffenberger)



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:34 CDT