Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!mit-eddie!genrad!decvax!tektronix!tekcrl!vice!tekfdi!videovax!stever
From: stever@videovax.UUCP
Newsgroups: comp.sys.amiga
Subject: Re: Some thoughts and questions about the Ranger
Message-ID: <4274@videovax.Tek.COM>
Date: Thu, 12-Mar-87 11:34:46 EST
Article-I.D.: videovax.4274
Posted: Thu Mar 12 11:34:46 1987
Date-Received: Sat, 14-Mar-87 00:52:18 EST
References: <140@tahoma.UUCP> <782@ssc-bee.ssc-vax.UUCP> <2748@jade.BERKELEY.EDU>
Reply-To: stever@videovax.Tek.COM (Steven E. Rice, P.E.)
Organization: Tektronix Television Systems, Beaverton, Oregon
Lines: 26
Keywords: 14MHz Blitter

In article <2748@jade.BERKELEY.EDU>, Mike Meyer (mwm@eris.BERKELEY.EDU)
writes:

> . . .

> Given that most blitter-type operations can be done in small loops,
> and the cache on the 68020 being large enough to hold small loops,
> this makes sense. After the first pass through the loop, it runs out
> of cache, so the time to do memory fetches/stores dominates.

An even bigger win is that if you have to process the data you are going
to blit, you can do so on the fly -- as you stream it through the 68020.
This saves one load and one store for each piece of data.

However, the blitter in the Amiga performs a very important function --
display refresh.  Because this is done in a way that doesn't interfere
with the CPU (if the CPU is executing out of FAST memory), both can run
at once.  So, how about _two_ 68020s -- one to run the programs and one
to perform all the display processing?  Then, use Video RAM as the
display memory, and . . .

					Steve Rice

----------------------------------------------------------------------------
new: stever@videovax.TV.TEK.COM
old: {decvax | hplabs | ihnp4 | uw-beaver}!tektronix!videovax!stever