[comp.sys.atari.st] Some errors

barrett@jhunix.HCF.JHU.EDU (Dan Barrett) (07/31/90)

	There was an article posted recently about the Atari TT, and it
contained some errors.  I wanted to point them out.  I'm sorry that I lost
the original poster's name; I believe this info was typed in from an official
Atari flyer in New Zealand.

>                 Comparison of standard features
>                    Amiga 3000     Mac IIcx       Atari TT030
>Clock speed         16 Mhz         -              32 Mhz

	Misleading -- the Amiga 3000 comes in 16 and 25 MHz models.

>RAM                 2 to 17        1 to 32        2 to 26

	Wrong -- the Amiga goes from 2-18 MB internally, but up to 1.7
	GIGABYTES of RAM is addressable as standard.

>  Max resolution    1280 x 480     Optional       1280 x 960
>                    Interlaced                    Non-interlaced

	This is a very misleading comparison.  The 1280x960 Atari mode is
	listed as "Moniterm monochrome."  If so, you need a Moniterm
	monitor... right?

	The Amiga also has such standard graphics modes.  They go up to
	1008x1024 resolution with a compatible monitor.

>  Max colours       32 Colours                    256 Colours
>                    from 4096                     from 4096

	Wrong --  Max colours is 4096 from a palette of 4096 on an Amiga.

>  Max video RAM     1 Mb                          8 Mb

	Wrong -- Max video RAM (called CHIP RAM on Amiga) is 2 MB on an
	Amiga.

>Expansion           Proprietary    NuBus          VME

	Only partially correct.  Amiga has both proprietary and IBM AT slots.
	I might mention that these "proprietary" slots are 32-bit DMA,
	running at the full speed of the bus/processor.  The name of the
	Amiga expansion type is "Zorro III".

	Another omission is that the Amiga has a processor slot for
	plugging in a 68040.  The machine is ready for it; just plug it in.

                                                        Dan

 //////////////////////////////////////\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
| Dan Barrett, Department of Computer Science      Johns Hopkins University |
| INTERNET:   barrett@cs.jhu.edu           |                                |
| COMPUSERVE: >internet:barrett@cs.jhu.edu | UUCP:   barrett@jhunix.UUCP    |
 \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\/////////////////////////////////////

baffoni@aludra.usc.edu (Juxtaposer) (08/01/90)

In article <5969@jhunix.HCF.JHU.EDU> barrett@jhunix.HCF.JHU.EDU (Dan Barrett) writes:
>
[stuff deleted]
>>Clock speed         16 Mhz         -              32 Mhz
>
>	Misleading -- the Amiga 3000 comes in 16 and 25 MHz models.
>
>>RAM                 2 to 17        1 to 32        2 to 26
>
>	Wrong -- the Amiga goes from 2-18 MB internally, but up to 1.7
>	GIGABYTES of RAM is addressable as standard.

   Certainly, due to the 200pins of the 68030, this is THEORETICALLY possible,
but due to the hardware constraints (limited card slots, bus width, power?:)
just how much can you ACTUALLY address in the REAL world (no flames, this is
an honest question)?

[stuff deleted]
>>  Max colours       32 Colours                    256 Colours
>>                    from 4096                     from 4096
>
>	Wrong --  Max colours is 4096 from a palette of 4096 on an Amiga.
>
	From what I understand, the 4k out of 4k colors obtained is due to 
software hacks (HAM methods, scanline interrupts, etc.), but I think that the 
question of comparing them is directed at standard system constraints rather 
than at what can be done by tweeking the normal draw cycles.
	So what would the max colours be WITHOUT scanline tweeking?

>
>                                                        Dan
>
> //////////////////////////////////////\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
>| Dan Barrett, Department of Computer Science      Johns Hopkins University |
>| INTERNET:   barrett@cs.jhu.edu           |                                |
>| COMPUSERVE: >internet:barrett@cs.jhu.edu | UUCP:   barrett@jhunix.UUCP    |
> \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\/////////////////////////////////////

-Mike

just Mike.
 

daveh@cbmvax.commodore.com (Dave Haynie) (08/08/90)

In article <11167@chaph.usc.edu> baffoni@aludra.usc.edu (Juxtaposer) writes:
>In article <5969@jhunix.HCF.JHU.EDU> barrett@jhunix.HCF.JHU.EDU (Dan Barrett) writes:

>[stuff deleted]
>>>Clock speed         16 Mhz         -              32 Mhz

>>	Misleading -- the Amiga 3000 comes in 16 and 25 MHz models.

>>>RAM                 2 to 17        1 to 32        2 to 26

>>	Wrong -- the Amiga goes from 2-18 MB internally, but up to 1.7
>>	GIGABYTES of RAM is addressable as standard.

>   Certainly, due to the 200pins of the 68030 

Not that it's revelent to this argument, but the PGA version of the 68030
has 128 pins, the surface mount version around 132 pins.

>this is THEORETICALLY possible,

Theoretically speaking, any 68030 system could have 4 GB of memory, just as
any 68000 system could have 16 MB of memory.  Practical considerations, such
as memory chunks reserved for I/O, OS, etc. will generally lower this number.
Physical realities may lower it some more.

>but due to the hardware constraints (limited card slots, bus width, power?:)
>just how much can you ACTUALLY address in the REAL world (no flames, this is
>an honest question)?

Today, as we speak, without building any additional hardware, I could plug
148 Megabytes of memory into my A3000.  Since no one else has access to the
32 Megabyte expansion card (though all Developers who made it to the Amiga
DevCon last June have the schematics), the practical limit this moment is
18 MB total.

Based on 4MB CMOS DRAM technology, it's certainly possible to build 64 MB
of memory onto a single Zorro III card.  It's also possible to fit at least
16MB on an A3000 Coprocessor Card. So based on today's available technology, 
the practical limit is 290MB.  This limit will of course grow as 16 or 64
megabit parts become available.  The A3000 reserves 128MB for memory 
devices that sit in the Coprocessor Slot, and 1.75GB for Zorro III devices.
This is, incidently, the first time Commodore has released a system that
will have to wait two memory generations before address space could again
be a limiting factor.  

Fortunately, humans aren't changing nearly as fast as computers.  Just about
everyone needs more than 64K of memory.  A great deal of folks have been 
satisfied with 1MB of memory (eg, the majority of PClone users).  Most everyone
will be satisfied with 16-18MB of memory.  The few who aren't should be happy
with 2-4 times that 16MB chunk for the moment.  Gradually, more folks will 
find a use for more than 16MB, but hardly in the volumes who found a use for
more than 64K or 1MB.  And this has all taken place in a single human 
generation; I can recall when I, personally, found a need for more than 16K
of memory.  Back in '79, 32K seemed practically unbounded.

>	From what I understand, the 4k out of 4k colors obtained is due to 
>software hacks (HAM methods, scanline interrupts, etc.), but I think that the 
>question of comparing them is directed at standard system constraints rather 
>than at what can be done by tweeking the normal draw cycles.

HAM, the 4K out of 4K they're undoubtedly referring to, is a standard hardware
display mode on the Amiga.  HAM does allow each pixel on the screen to display
a different color, however, not an arbitrary color.  The color of each 
successive pixel is based on either 1 of 16 random colors from the CLUT, or
a R, B, or G modification of the color from the previous pixel.  

There are apparently "dynamic modes", or hacks, used on the Amiga for more
colors in other standard display modes.  Some simple ones may use the display 
coprocessor to change colors, though I suspect the more advanced ones use a 
display interrupt and have the CPU bang the color registers.  In any case, I
suspect the same kind of things can be done on other systems with reasonable
display hardware (even the C64 had a primitive scan line interrupt), so these
must certainly NOT be considered standard display modes.  A good comparison
between systems might indicate that the display hardware has additional 
features, such as scan-line interrupts, a display coprocessor, or a 
bimmer/blitter.

>>                                                        Dan

>-Mike


-- 
Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests"
   {uunet|pyramid|rutgers}!cbmvax!daveh      PLINK: hazy     BIX: hazy
      Get that coffee outta my face, put a Margarita in its place!