Amiga 3000UX, X, OpenLook, Motif, Color, A2410, Etc. (somewhat long)

[Randell Jesup]

In article <1991Mar30.090353.9749 at kessner.denver.co.us> david at kessner.denver.co.us (David D. Kessner) writes:
>Here, I take an assumption that the 386 is faster than the 030.  For the
>sake of that paragraph IT DOESNT MATTER.  What I am saying here is that the
>030 is a cleaner design than the 386.  Given the age of the Motorola line
>of CPU's (in comparison to the age of Intel's _32_bit_ CPU's) I'm suprised
>that the 030 doesnt blow the socks off the 386 (by that I mean 2-3 times
>faster, which it is clearly not).

	I wouldn't expect it to when the 386 isn't in MSDOS mode.  Effectively,
the 386 is a faster processor with a 286 in a corner (yeah, I know, it's
not _really_ like that, though the 486 is fairly close to that).

	In general, Moto and Intel CPUs of the same general vintage are not
too far apart (they've been getting closer: 68000 was about equal to 6Mhz
80286 - better in some cases, slower in other).  Witness the '486 and the
'040: last I saw, the '040 gets 11.8 specmarks, and the 486 gets 8.8.  However,
for integer operations (dhrystone is integer, though weighted to unusual
strings), the '040 gets 12.9, and the '486 gets 13.3 (the '040 FP is 11.0 to
6.6 for the '486).  Note that those were both with 128K of external cache,
at 25Mhz.

>need Dhrystone 2.x, the 386 is just faster, etc.  I personally tend to believe
>the lack of cache, and 'string based 386' myself.  The other reasons could be
>valid, so I am keeping my options open.

	If those 386 results are from a cached machine, I'm certain the '386
would be faster on dhrystone.  dhrystone is a semi-cachebuster on an '030,
but most of it fits easily in even a small external cache (4K/4K).

>Benchmarks ran under MS-DOS are misleading at best.  Often the results are
>half the speed as what is achieved under UNIX.  My machine runs Dhrystones
>at 7500 under MS-DOS/Turbo-C++, but 11,900 under UNIX/GCC.  This is due to
>the lack of 32 bit regesters/optimization, having to deal with 64K segments,
>and the early (pre-387) FPU's required the use of the FWAIT before each 
>FPU instruction (it pauses the CPU till the FPU is done) and this slows
>things down dramatically.  In addition, the 387 has trig instructions while
>MSC (microsoft C for those slow folks) does not utilize this.

	Note that dhrystone has no FP stuff in it at all.  Specmarks are FAR
better for any sort of discussion like this.  In 680x0 code, "small-model"
code is usually faster/smaller than 32-bit code (though of course on a 680x0
you're not forced into either).  This is one reason why AmigaDos compilers
can produce better numbers than the Unix compilers do.  Note the performance
of the '486 on FP stuff under Unix in the SpecMarks above, also.

>	386/33 with 34010 board for about $6000.
>	486/33 for about $7000
>	486/33 EISA for about $8000
>	SPARC Clones for about $8000 <-- includes 16" color monitor.
>	and the Color NeXT <-- 17" monitor, 16 bits/pixel

	One thing to remember to add: disk and unix itself.  Amiga Unix has
a 200Meg disk.  I'm not certain things like diskless Sparcs include Unix
tapes.

-- 
Randell Jesup, Keeper of AmigaDos, Commodore Engineering.
{uunet|rutgers}!cbmvax!jesup, jesup at cbmvax.commodore.com  BIX: rjesup  
Disclaimer: Nothing I say is anything other than my personal opinion.
Thus spake the Master Ninjei: "To program a million-line operating system
is easy, to change a man's temperament is more difficult."
(From "The Zen of Programming")  ;-)