Prior Art, RAS/MMU overlap, RAS from low address lines

John Bass bass at dmsd.UUCP
Tue Sep 23 14:43:31 AEST 1986


It would be nice if SUN would post exactly what they claim to patent,
or else someone who has the time/interest could visit the Patent Library
in the Bay area to get and post a copy of the patent description.
The SUN readers have been amazingly quite thru this whole discussion,
did they get a gag order from management?

I do know that the Onix C8002 (designed 1979, shipped 1980), Tandy Model 16
(designed 1981 (1982?), shipped 1982), and Fortune 32:16 (designed 1981,
shipped 1982) all use RAS from the low address lines. The Tandy and Fortune
seem to overlap mmu translation with ras prior to cas, it is unclear what the
ONIX does. All three systems have published schematics and hardware reference
manuals, Tandy's can be purchased over the counter without any trouble, Fortune
requires a developers agreement, and I don't know what the status of the ONIX
manual is now days -- it used to be available with little trouble.

The Onix mmu is based on mapping rams with 2k pages, the Tandy and Fortune use
base/limit mmu's.

I know of several multiuser Z80 designs with bank switched memory that use
nearly identical technology (RAS from low address lines, CAS from high address
lines and bank register) all dating back to the mid/late 70's. These could
hardly claim to overlap anything since the bank address is a static latch.
Anybody know of a design based on mapping rams or register files?

Unless SUN is claiming much more than RAS/CAS overlap, it would be nice if
they abandoned the patent voluntarily and save everyone the legal fees
to formally contest and defeat it. How about it you guys at SUN?? Do you
really claim more than this??
-- 

John Bass (DBA:DMS Design)
DMS Design (System Design, Performance and Arch Consultants)
{amdahl,fortune,polyslo}!dmsd!bass



More information about the Comp.unix.wizards mailing list