Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wuarchive!uunet!munnari.oz.au!bunyip.cc.uq.oz.au!marlin.jcu.edu.au!cpca
From: cpca@marlin.jcu.edu.au (Colin Adams)
Newsgroups: comp.sys.amiga.advocacy
Subject: Re: (Video) Hardware Idiots ?
Message-ID: <1991Jun13.042836.8112@marlin.jcu.edu.au>
Date: 13 Jun 91 04:28:36 GMT
References: <1991Jun9.060440.29078@leland.Stanford.EDU> <1991Jun10.065629.21255@marlin.jcu.edu.au> <22362@cbmvax.commodore.com>
Organization: James Cook University of North Queensland
Lines: 27

In article <22362@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes:
>In article <1991Jun10.065629.21255@marlin.jcu.edu.au> cpca@marlin.jcu.edu.au (Colin Adams) writes:
>>I told him to try using CopyMemQuick and he got a 75% increase in speed.
>>Nough said!
>
>Not really.  You have done a simple test of the best case for the 68030 vs.
>the blitter.  Now try the worst case.  Instead of longword aligned blocks of
>memory, try a situation where source and destination are on different bit
>boundaries.  Next, try a three operand blit operation, where all three operands
>are on different bit boundaries.  You'll see that the blitter does the 
>nontrivial cases pretty well.

This is probably due to the inbuilt barrel-shifter in the blitter, I 
think the 030 has one too.  If anyone has done this test, I'd be
interested (and I'm sure others would be too) to see how the blitter
compares.

>Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests"
>   {uunet|pyramid|rutgers}!cbmvax!daveh      PLINK: hazy     BIX: hazy
>	"This is my mistake.  Let me make it good." -R.E.M.


-- 
Colin Adams                                  
Computer Science Department                     James Cook University 
Internet : cpca@marlin.jcu.edu.au               North Queensland
'And on the eighth day, God created Manchester'