Subject: Re: NS638 RAM
To: None <port-vax@netbsd.org>
From: Carl Lowenstein <cdl@mpl.ucsd.edu>
List: port-vax
Date: 03/24/2003 23:11:21
> From: der Mouse <mouse@Rodents.Montreal.QC.CA>
> Date: Mon, 24 Mar 2003 18:44:26 -0500 (EST)
> To: port-vax@netbsd.org
> Subject: Re: NS638 RAM
> 
> > I wonder what the practical limitation would be on the number of
> > KA630s in such a system.  I've been doing a lot of assembler lately,
> > and I have a large number of KA630s and I'm having very sick
> > thoughts. :)
> 
> There are only two bits of ID, so at most four processors per Qbus.
> 
> Unless you're willing to live with just the on-board 1M, you will also
> want memory boards, which call for CD-interconnect slots.  The most
> Q/CD slots I know of occur in the BA123[%], with four, which would
> limit you to two CPUs if each is to have at least one board of RAM.

There were LSI-11 backplanes with nothing but CD-interconnect slots,
9 slots as I recall.  Qbus on AB, interconnect on CD.
11/23 systems came that way at one time.

They might not have been wired for 22-bit addressing on the Qbus,
but that is easy to fix.

> 
> [%] Do I have the BA numbers right?  I can't find any BAnnn marking on
>     mine....

Far as I can tell, yes.

    carl
-- 
    carl lowenstein         marine physical lab     u.c. san diego
                                                 clowenst@ucsd.edu