From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <3cb7441037fe3049b0c6f5794f8fd4a8@mightycheese.com>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Gnot
From: "rob pike, esq." <rob@mightycheese.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit
Date: Fri, 25 Oct 2002 08:00:53 -0400
Topicbox-Message-UUID: 0d7ba2ce-eacb-11e9-9e20-41e7f4b1d025

the blit work was a collaboration between bart locanthi and myself,
as was the gnot. the gnot used the case and monitor from the 630
but was all new inside. it had a 68020. the 'crisp gnot' came later
and didn't really succeed due to various chip problems. the 68020
version was widely deployed and pretty successful within the
research group at bell labs.

one of the new pieces of the crisp gnot was a special support chip
for graphics, to compensate for the poorer bit swizzling capability
of the crisp relative to the 68020.  the chip watched the bus and
did things to memory as it flew by.  it was a pretty clever idea but
failed because the crisp would double-execute stores and fetches
across page faults, causing the BALU to double-execute the graphics
ops.

-rob