KA620

From: Michael Sokolov <msokolov_at_ivan.Harhan.ORG>
Date: Fri Mar 19 12:28:28 2004

der Mouse <mouse_at_rodents.montreal.qc.ca> wrote:

> Possibly - or perhaps a KA620 is one example of an rtVAX, much as a
> KA630 is an example of a VAX.

You are right of course conceptually. But I'm pretty sure now that KA620 was
the only implementation of the rtVAX architecture.

> My VARM is "Revision 6.1" and is not obviously marked with an edition
> number.

1982-05-20, right? I have that one too, it's my first VARM, one of Tim Shoppa's
early gifts to Quasijarus Project. The early VARMs, the EK-VAXAR-RM ones, are
lovely: 100% pure ASCII, except for boldface section headers the book is line
printer output. From the appearance seems to be very close to the DEC internal
VAX spec, DEC STD 032 (though not having a copy of the latter it's an educated
guess). But then they stopped issuing VARMs in that format, and published it as
a book (with ISBN and all, rather than an EK part #). The book version of VARM
had two editions: 1st ed. in 1987 by Tim Leonard (listed as the keeper of the
DEC internal VAX spec in an appendix to another highly secretive DEC internal
spec I have) and 2nd ed. in 1991 by Richard Brunner (wonder what happened to
Tim...). I have acquired both of these books recently. Not having a copy of
the genuine article (the DEC internal STD 032 spec) I have to live with the
published versions instead, so I was trying to acquire as many of them as
possible to complete my mental picture of the complete architecture spec and its
variations. (I'm designing a new VAX CPU chip, so I have to have a very solid
picture of the spec requirements and options and their evolution.)

Would anyone perchance have a copy of the real VAX spec, DEC STD 032 aka
EL-00032-00 aka A-DS-EL00032-00-0?

> I don't recall seeing an rtVAX mentioned in it, and the index
> does not list anything beginning with rt-.

Yeah, of course the 1982-05-20 VARM predates it by a few years.

Brian Chase <vaxzilla_at_jarai.org> wrote:

> The 2nd Ed VARM has a section on rtVAX memory management, 11.2.3 on pgs
> 422-424. I don't yet see any other mentions of it the book to indicate
> whether it actually is designated as the KA620.

Table 8-2 on pp. 331-332 lists the SID code assignments, and 16 decimal is
listed as "rt/uVAX (chip 78R32)". Knowing that 78032 is the MicroVAX II chip,
used in KA630, KA410 (MV/VS2000), and a bunch of other less-known systems (see
the SYS_TYPE codes the same table lists for SID 08), it's pretty obvious that
the rtVAX 1000 listed by the table as being based on this 78R32 is the KA620.

Since the MMU is internal to the CPU chip, the difference between KA630 and
KA620 is in the silicon (78032 vs. 78R32) rather than board-level.

MS
Received on Fri Mar 19 2004 - 12:28:28 GMT