[comp.sys.ibm.pc] Help! Where's the lost 128K memory?

cwang@copper.ucs.indiana.edu (Ted Wang) (06/09/91)

Hello!  I have a 80386-25 based PC with 64K cached memory and 4MB RAM.
It uses a noname motherboard and AMI BIOS.  When I turn on the
computer, it does self memory testing and says 3968 OK instead of 4096
OK.  I thought that this might because of the SIMMs, so I pulled out
the SIMMs ( four 1MBx9 SIMMs) in my computer and plugged into my
friend's computer and everything's fine (I mean he got 4096 OK message
after the memory test).  Any idea of where the lost 128K goes in my
computer?  Thanks in advance!

Ted Wang
cwang@copper.ucs.indiana.edu

shutton@bronze.ucs.indiana.edu (Scott K. Hutton) (06/09/91)

cwang@copper.ucs.indiana.edu (Ted Wang) writes:
>Hello!  I have a 80386-25 based PC with 64K cached memory and 4MB RAM.
>It uses a noname motherboard and AMI BIOS.  When I turn on the
>computer, it does self memory testing and says 3968 OK instead of 4096
>OK.  I thought that this might because of the SIMMs, so I pulled out
>the SIMMs ( four 1MBx9 SIMMs) in my computer and plugged into my
>friend's computer and everything's fine (I mean he got 4096 OK message
>after the memory test).  Any idea of where the lost 128K goes in my
>computer?  Thanks in advance!

There  was a question with  a similar answer earlier.   If you look in
your CMOS setup, you will probably see that you have either or both of
your Main and Video BIOS shadowed.  This will automatically take about
192K out before you get to the memory  count.   Works  on mine!  Don't
worry... it's still there; it's just being used!

  	
 _____________________________________________________________________________
  Scott K. Hutton, KA9WTR           Internet:  SHutton@UCS.Indiana.edu
  UCS Support Services                         SHutton@Bronze.UCS.Indiana.edu
  Indiana University                  Bitnet:  SHutton@IUIS, SHutton@IUBACS

ckp@grebyn.com (Checkpoint Technologies) (06/09/91)

In article <1991Jun8.201349.14697@bronze.ucs.indiana.edu> cwang@copper.ucs.indiana.edu (Ted Wang) writes:
>Hello!  I have a 80386-25 based PC with 64K cached memory and 4MB RAM.
>It uses a noname motherboard and AMI BIOS.  When I turn on the
>computer, it does self memory testing and says 3968 OK instead of 4096
>OK.

(Sorry, this is not an answer to your question Ted...)

A similar thing happens to me, but stranger.  With 4M installed, the
self test says 4096K, which is what I expect.  But with 8M installed, it
says 8064K - which is 128K short. I have a 386-33, apparently made by a
company called TTL, and an AMI BIOS.
-- 
Richard Krehbiel, private citizen      ckp@grebyn.com
(Who needs a fancy .signature?)

exnirad@brolga.cc.uq.oz.au (Nirad Sharma) (06/10/91)

ckp@grebyn.com (Checkpoint Technologies) writes:

>In article <1991Jun8.201349.14697@bronze.ucs.indiana.edu> cwang@copper.ucs.indiana.edu (Ted Wang) writes:
>>Hello!  I have a 80386-25 based PC with 64K cached memory and 4MB RAM.
>>It uses a noname motherboard and AMI BIOS.  When I turn on the
>>computer, it does self memory testing and says 3968 OK instead of 4096
>>OK.

>(Sorry, this is not an answer to your question Ted...)

>A similar thing happens to me, but stranger.  With 4M installed, the
>self test says 4096K, which is what I expect.  But with 8M installed, it
>says 8064K - which is 128K short. I have a 386-33, apparently made by a
>company called TTL, and an AMI BIOS.
>-- 
>Richard Krehbiel, private citizen      ckp@grebyn.com
>(Who needs a fancy .signature?)

I don't know if this will help but when I had exactly the same problem with
an IBM PS/2 Model 80 I was told that as soon as installed memory exceeds a
certain level (2M for me) the system starts ROM shadowing.  This could be
what's happening to you people.  Shadowing isn't much use to me as I run 
SCO Xenix which ignores most of the system ROM.
-- 
Nirad Sharma  (exnirad@brolga.cc.uq.oz.au)		Phone : (+61 7) 365 7575
Systems Programmer					Fax :	(+61 7) 870 5080
Continuing Education Unit
The University of Queensland.  QLD  4072

jpollard@silver.ucs.indiana.edu (06/13/91)

In article <1991Jun8.201349.14697@bronze.ucs.indiana.edu> cwang@copper.ucs.indiana.edu (Ted Wang) writes:
>Hello!  I have a 80386-25 based PC with 64K cached memory and 4MB RAM.
>It uses a noname motherboard and AMI BIOS.  When I turn on the
>computer, it does self memory testing and says 3968 OK instead of 4096
>OK.  I thought that this might because of the SIMMs, so I pulled out
>the SIMMs ( four 1MBx9 SIMMs) in my computer and plugged into my
>friend's computer and everything's fine (I mean he got 4096 OK message
>after the memory test).  Any idea of where the lost 128K goes in my
>computer?  Thanks in advance!
>
>Ted Wang
>cwang@copper.ucs.indiana.edu
>

brigc@world.std.com (Brig C McCoy) (06/13/91)

In article <1991Jun12.192458.16961@bronze.ucs.indiana.edu> jpollard@silver.ucs.indiana.edu () writes:
>>Hello!  I have a 80386-25 based PC with 64K cached memory and 4MB RAM.
>>It uses a noname motherboard and AMI BIOS.  When I turn on the
>>computer, it does self memory testing and says 3968 OK instead of 4096
  [...  ...]
>>Any idea of where the lost 128K goes in my computer?  Thanks in advance!
>>
>>Ted Wang
>>cwang@copper.ucs.indiana.edu

I'm not positive about this particular piece of hardware, but I know several
of the newer machines are set up to use 128K to "shadow" the ROM code... that
is copy the ROM code to the 128K of RAM where it'll run faster.

+-----------------------------------------------------------------+
| Brig C. McCoy                 --  "Standard disclaimers"        |
| Systems Administrator         --                                |
| Soft-tek International        --  Internet: brigc@world.std.com |
| 1999 North Amidon, Suite 340  --     Voice: 316 838-7200        |
| Wichita, KS 67203-2124        --       Fax: 316 838-3789        |
+-----------------------------------------------------------------+

waltp@bingvaxu.cc.binghamton.edu (walter piotrowski) (06/13/91)

The missing 128kB holds a RAM copy of the ROM BIOS.  It's copied
into RAM for speed.

-- 
-----------------------------------------------
Walter G. Piotrowski
Computer Science Department - Watson School
State University of New York