Idea for a graphics board

105 views
Skip to first unread message

Charles Cazabon (186-003-526)

unread,
May 28, 1991, 9:46:20 PM5/28/91
to
In article <1991May28.1...@nntp-server.caltech.edu> nyg...@nntp-server.caltech.edu (Michael T. Nygard) writes:
>Here's a description of my dream graphics board
>
>i860 processor for anti-aliasing lines, fast 3d polygon renders, Goroud & Phong
>shading. (64 bit busses.)
>4M memory for frame buffer & z-buffer
>1024x1024x24 resolution
>8 bit z-buffer
>
>Anybody know of such a beast, or anything similar. If not, I'll start doing
>some designs.
>

>Replies & suggestions are welcome
Replies and suggestions welcome...my suggestion is, go for it!
You sound like you know what you are doing...I sure hope so. If so, can you
forcast what ballpark price it would sell for? I realize it would be a real
"guesstimate", but say, is it <$1000, $1000-$2000, or >$2000?

This is the perfect thing for me...if you find that the i860 processor is
either too expensive to implement, or does not seem too well suited for hardware
antialiasing, or such, perhaps you could see what is on the ATI VGAwonder
board. Evidently it does hardware antialiasing as well.

My suggestions for the product:

Support for other resolutions than 1024x1024. Say, the common ones, like
320x200, 320x400, 640x200, 640x400, 1024x768, etc. Oh, plus overscan would
be nice.

Also, other colour resolutions. Say, be able to kick it down into 8 bit colour
or 8 bit greyscale modes.

If it is not desgined for the video slot, how about making it Zorro III, so
us 3000 owners (I think you said you have one as well) can take advantage
of 32 bit bussing?

I'm too excited as it is--don't let this product be vapourware. Oh yes, I have
yet to hear of such a "dream card" for the Amiga.

-Chuck
caz...@hercules.uregina.ca

Michael T. Nygard

unread,
May 28, 1991, 3:10:19 PM5/28/91
to
Here's a description of my dream graphics board

i860 processor for anti-aliasing lines, fast 3d polygon renders, Goroud & Phong
shading. (64 bit busses.)
4M memory for frame buffer & z-buffer
1024x1024x24 resolution
8 bit z-buffer

Anybody know of such a beast, or anything similar. If not, I'll start doing
some designs.

Replies & suggestions are welcome!

--
--------
Mike Nygard
nyg...@coil.cco.caltech.edu

Clayton Cowgill

unread,
May 29, 1991, 11:09:03 PM5/29/91
to
In <1991May29.2...@nntp-server.caltech.edu> nyg...@nntp-server.caltech.edu (Michael T. Nygard) writes:

>I've been looking at some numbers on the board. Here's what I came up with.

>Without using a CLUT (yuk.), assuming 80ns RAM, 1024x1024x24 resolution.

>1/(1024*1024*24*60) = .66227 ns/bit.
>(80 ns/cycle) / (.66227 ns/bit) = 120.8 bits/cycle

>That means a bus _at least_ 128 bits wide. A 128 bit bus results in 5% of
>the memory cycles availible for the processor.

>Unless I go to video RAM, things could get unreasonable very fast.
>I know little about video RAM, could someone fill me in?

You might look at the TI34020 and its new FPU co-processor. It provides a lot
of the functionality of the i860, plus the posibility of TIGA compatibility.
The 34020 also has a 'smart' memory addressing system that can be used with
TI VRAMs- it radically improves memory access throughput.

4Meg sounds pretty thin for the kind of resolutions you're talking about.
Monotors for a mega-pixel display are pretty expensive right now too... With
a little luck XGA will catch on in the PC world and monitor prices will come
down...

The TI Graphics FPU (34082) has a 32MHz clock and will perform IEEE single
precision divide ops in 8 clock cycles and square root in 11 cycles. There
are 30 2-d and 3-d transforms in hardware with goodies like 1x4 and 4x4
matrix operations and clipping. (for review, the 34020 allows 512Mbytes of
addressing, pixel processing, programmable video rates, etc...)

'Eh... Just my $.02

-Clay

##################===============------------===============##################
Clayton Cowgill cow...@prism.cs.orst.edu
Head Consultant Lab: (503) 737-2435
Oregon State University CS Dept. Home: (503) 757-7060
Corvallis, OR 97330

F M Cargill

unread,
May 29, 1991, 3:22:36 PM5/29/91
to
In article <4...@regina.uregina.ca> caz...@hercules.uregina.ca (Charles Cazabon (186-003-526)) writes:
>In article <1991May28.1...@nntp-server.caltech.edu> nyg...@nntp-server.caltech.edu (Michael T. Nygard) writes:
>>Here's a description of my dream graphics board
>>
>>i860 processor for anti-aliasing lines, fast 3d polygon renders, Goroud & Phong
>>shading. (64 bit busses.)
>>4M memory for frame buffer & z-buffer
>>1024x1024x24 resolution
>>8 bit z-buffer
>>
>>Replies & suggestions are welcome
>My suggestions for the product:
>
>Support for other resolutions than 1024x1024. Say, the common ones, like
>320x200, 320x400, 640x200, 640x400, 1024x768, etc. Oh, plus overscan would
>be nice.
yeah I'll go for that - you look like you want to do a graphics processor
card rather than a frame buffer. If so you'll want to have the capacity
to add more bitplanes - an 8bit Z buffer is too limiting.
You might also consider including hardware Z buffer projections, just so we
can render a scene at 60 frames/sec. I guess that comes under Gouraud & Phong
shading...

As for antialiased lines, these are quite easy - you might as well include
conic sections which are also easy to do in hardware and antialiasing on them
too. These things mean you don't have to use caches for your outline fonts -
you just draw 'em where you need 'em.

>I'm too excited as it is--don't let this product be vapourware. Oh yes, I have
>yet to hear of such a "dream card" for the Amiga.

But wait, where's the market for it? Well if you make it so that it'll do 3D
stuff in hardware then it'll knock the socks off an '040 and anybody who
wants to do animations in 3D will be knocking down your door in a rush to get
one.

As for existing products there was a post a couple of weeks back of a card
that used Texas Instruments' graphics processors - I understand that it
interfaces to the 680X0s better than the i860?

Yet another vertical market opening up to the Amiga?

--
* Fletch Cargill cs8...@cc.brunel.ac.uk * // *
* Brunel University Department of Computer Science * \X/ *

Michael T. Nygard

unread,
May 29, 1991, 5:23:15 PM5/29/91
to
I've been looking at some numbers on the board. Here's what I came up with.

Without using a CLUT (yuk.), assuming 80ns RAM, 1024x1024x24 resolution.

1/(1024*1024*24*60) = .66227 ns/bit.
(80 ns/cycle) / (.66227 ns/bit) = 120.8 bits/cycle

That means a bus _at least_ 128 bits wide. A 128 bit bus results in 5% of
the memory cycles availible for the processor.

Unless I go to video RAM, things could get unreasonable very fast.
I know little about video RAM, could someone fill me in?

Other problems:
I hadn't planned on this ever being a commercial product. Getting the kind
of printed circuit board this would require in single quantity would be very
expensive.

As for the cost estimate, it's hard to say. Could be real high, maybe not.
There are too many factors at this stage (early design stage). No less than
$1000 for a one-of-a-kind. Now if we could put together a few people...

This probably belongs in .hardware, but it started here, and I think the
main interest is here.

Free free to contact me if you have suggestions (or would like to help!)

Andre T. Yew

unread,
May 29, 1991, 9:06:21 PM5/29/91
to
nyg...@nntp-server.caltech.edu (Michael T. Nygard) writes:

>Here's a description of my dream graphics board

>i860 processor for anti-aliasing lines, fast 3d polygon renders, Goroud & Phong
>shading. (64 bit busses.)
>4M memory for frame buffer & z-buffer
>1024x1024x24 resolution
>8 bit z-buffer

Since we're talking dream graphics boards here, why don't we go
all out? How about a pipelined system, a la Silicon Graphics? I guess
we could have several i860's pipelined together. The first one might do
the transformations (rotation, translation, scaling, figuring out the
surface normals). The second pipeline could do all the lighting. The
third could do clipping. The fourth perspective transformations, and
the fifth just converting everything to device coordinates. Since you don't
want to bog down the pipeline too much, some tasks that might be bottlenecks
could be spread out over several processors. For those who are wondering,
yes this sequence is shamelessly ripped out of the SGI GTX pipeline. I think
that if you do it this way, the limiting factor would be how fast you could
move your data onto your screen, especially since we'll have just one 860
doing the scan-conversion and Gouraud interpolation (yes, I'm interpolating
colors in device coords.). Of course, to control this pipeline, each i860
would have to have some kind of input FIFO buffer that's fast enough to
grab things from the previous processor and some sort of local memory for
scratch space and maybe storage if we get a pipeline stall. The way the SGI
GTX does the raster part of the system is to fan out to 20 Image Engines.
Each engine controls 1/20th of the screen, or 64K pixels on a 1280x1024
screen. Perhaps we could have some other processor that talks well to
an 860 to handle this part instead of that last i860. My guess is that we
should have some processor that can address lots of memory (21 bits or more
of address space, depending on what special effects you'd like to have) and can
DMA very fast in and out of that memory. How about an i960CA? If you want
to do some special effects, you'd have to have some computation power, too.

Now, for the specs. Let's say we're doing a 1280x1024 screen and we
have 20 screen processors. Let's say they're 960's, but we decide to put
only 2^24 bits of memory on them (16MB). That means that each pixel would
get (each 960 controls 2^16 pixels) 2^8 bits. That's 256 bits per pixel.
You could divide this up in to two 24 bit buffers for double-buffering, with
8 bits per buffer for alpha channels and you have 192 bits left. How about
12 bits for a windowing system, and 24 bits for a Z-buffer. The remaining
156 bits can be used for texture maps or other special effects. I have
absolutely no idea what the throughput of the pipeline would be.

So there's my dream graphics board, or at least a gross caricature
of it. Memory would probably be as expensive at least as the processors.
I figure an upper limit of 2^25.5 bits of memory. It would be nice too to have
some way of writing over the code for each stage of the pipeline so we could
use it for some other things too. Perhaps we could attach ROMs to each i860
with the normal graphics code that they could copy into their own, say, SRAM
whenever we initialize the pipeline for graphics. And when we want to,
we could have the i860 load its programs from somewhere else so we could
write our own applications for the pipeline. BTW, programming this thing
would probably be very difficult.

Andre

--
Andre Yew and...@through.cs.caltech.edu (131.215.131.169)

Tom Limoncelli +1 201 408 5389

unread,
May 29, 1991, 11:15:12 PM5/29/91
to

> Here's a description of my dream graphics board
>
> i860 processor for anti-aliasing lines, fast 3d polygon renders, Goroud & Phong
> shading. (64 bit busses.)
> 4M memory for frame buffer & z-buffer
> 1024x1024x24 resolution
> 8 bit z-buffer

Why?

I tell you what. I'll design it if you pay for it. That's right,
just sending me checks and you'll have it in a reasonable amount of time.
Until I start seeing those checks, please take this to
comp.sys.amiga.advocacy where it belongs.

Hrrrumph.

Tom
--
Tom Limoncelli tlim...@drew.edu tlim...@drew.bitnet 201-408-5389
"People in tight pants, moving fast."

Dave Haynie

unread,
May 30, 1991, 2:11:29 AM5/30/91
to
In article <1991May29.2...@nntp-server.caltech.edu> nyg...@nntp-server.caltech.edu (Michael T. Nygard) writes:
>I've been looking at some numbers on the board. Here's what I came up with.

>Without using a CLUT (yuk.), assuming 80ns RAM, 1024x1024x24 resolution.

>1/(1024*1024*24*60) = .66227 ns/bit.
>(80 ns/cycle) / (.66227 ns/bit) = 120.8 bits/cycle

>That means a bus _at least_ 128 bits wide. A 128 bit bus results in 5% of
>the memory cycles availible for the processor.

Uh, not to burst your bubble or anything, but the 80ns in an 80ns RAM chip
is the Row Access Time, not the cycle time. The cycle time is usually
around 150-160ns. Of course, with page or static column mode DRAM, an 80ns
part may do an on-page cycle in around 50ns.

>Unless I go to video RAM, things could get unreasonable very fast.
>I know little about video RAM, could someone fill me in?

VRAM is likely what ya want on a board like that. A VRAM has two ports, the
normal RAS/CAS random access port, and a serial shift register port. You can
run a wacky cycle that, rather than pull a page or static mode trick, will
download a whole row into the shift register. You can now proceed to shift
that data out quickly, while at the same time using the random access port for
CPU/GPU/whatever things.

>Mike Nygard


--
Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests"
{uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy
"That's me in the corner, that's me in the spotlight" -R.E.M.

Charles Cazabon (186-003-526)

unread,
May 30, 1991, 5:17:07 AM5/30/91
to
In article <1991May29.2...@nntp-server.caltech.edu> nyg...@nntp-server.caltech.edu (Michael T. Nygard) writes:
:I've been looking at some numbers on the board. Here's what I came up with.

:
:Without using a CLUT (yuk.), assuming 80ns RAM, 1024x1024x24 resolution.
:
:1/(1024*1024*24*60) = .66227 ns/bit.
:(80 ns/cycle) / (.66227 ns/bit) = 120.8 bits/cycle
:
:That means a bus _at least_ 128 bits wide. A 128 bit bus results in 5% of
:the memory cycles availible for the processor.

I know that you can get 60ns ram pretty cheap (almost as cheap as 80ns), and
there are faster ram out there for a little more money (40ns, for instance).
You could probably cut the bus down to 64 bits this way. Colourburst uses
a 48 bit bus if I am not mistaken. I think it would be easier to have your
own memory on board and let the graphics processor handle it at max speed.
Perhaps something on the card to do DMA in a blitter-like way?

:Unless I go to video RAM, things could get unreasonable very fast.


:I know little about video RAM, could someone fill me in?
:
:Other problems:
:I hadn't planned on this ever being a commercial product. Getting the kind
:of printed circuit board this would require in single quantity would be very
:expensive.
:
:As for the cost estimate, it's hard to say. Could be real high, maybe not.
:There are too many factors at this stage (early design stage). No less than
:$1000 for a one-of-a-kind. Now if we could put together a few people...

Not one of a kind. To lay out the circuit board, and have 20 printed up,
would probably only cost twice as much as a single board. Also, you could
get parts a lot cheaper in 20x quantities. Perhaps you should take orders
over the net, ask for a deposit, once you have a breadboard somewhat working.
Or at least a semi-complete design. I have a high-tech friend who could
field answers for you on just about any electronic/video/computer subject.
Just email me the questions at caz...@hercules.cc.uregina.ca (or is it
caz...@hercules.uregina.ca?). I will pass them on...he is great at
designing his own stuff. Perhaps if you did a schematic, he could lay out
the circuit board for you. He also works reasonably cheaply.

This could be the beginning of a new era. You know there is a market for
a $1000 or $1500 board of this type. I figure you could sell 1,000
boards in the first year. If you made $300 a board, you could be a rich
man, and quick!

I will gladly help out in any way I can...research on obscure topics, parts
sourcing, etc. Contact me for details (I work free for the good of the
Amiga community).


--Chuck Cazabon, caz...@hercules.cc.uregina.ca
* My Opinions Are Not My Own...Feel Free To Plagiarize

John Schultz

unread,
May 30, 1991, 2:25:12 PM5/30/91
to

>In article <1991May29.2...@nntp-server.caltech.edu> nyg...@nntp-server.caltech.edu (Michael T. Nygard) writes:
>>I've been looking at some numbers on the board. Here's what I came up with.

>>Without using a CLUT (yuk.), assuming 80ns RAM, 1024x1024x24 resolution.

>>1/(1024*1024*24*60) = .66227 ns/bit.
>>(80 ns/cycle) / (.66227 ns/bit) = 120.8 bits/cycle

[stuff deleted]

>VRAM is likely what ya want on a board like that. A VRAM has two ports, the
>normal RAS/CAS random access port, and a serial shift register port. You can
>run a wacky cycle that, rather than pull a page or static mode trick, will
>download a whole row into the shift register. You can now proceed to shift
>that data out quickly, while at the same time using the random access port for
>CPU/GPU/whatever things.

Yes, VRAM is a must. You also have to find a suitable RAMDAC (similar to
a Brooktree Bt473). Further, you'll need to communicate between the i860
and the host. For maximum throughput you'll need dual port ram accessible
by the Amiga and the i860. The 64-bit external data bus will also make things
even more expensive (if you want to utilize the full power here). The 7Mhz
16-bit bus of Zorro-II won't do for speed- Zorro-III is a must (The market
will be limited to A3000s and its successors).
Designing and building the hardware is the easy part (debugging it is a
little more work). Once the hardware is debugged, you'll need to write
software to utilize the enormous power of the i860. Although there are
C and Fortran compilers for the i860, you'll need to write your code
in RISC assembly to take complete advantage of the pipelining and paralellism.
If you don't, the code won't be much faster than if it were run on a
MIPS R3000 type processor. If you loved 680x0/80x86 assembly and enjoy getting
down to the nitty gritty, and writing a page of code to do a divide, then
you'll love the i860 (You'll especially enjoy writing and debugging pipelined
parallel code. You also get to play with the caches, and can even switch
between big and little endian formats).
The i860 also has built-in RISC instructions to do Gouraud shading with
16-32 bit (pipelined or not) z-buffering. See page 9-21 of the i860 programmer's
reference manual for more details and source code examples of 3-D rendering
techniques. The Intel blurb sheets claims 40,000 Grouraud/Z-buffered polygons
per second. That's 1333 polygons per frame in a system running at 30 frames/sec.
For graphics work, 3-D libraries have already been written. I viewed a demo
at HP-Appollo where a 68030 Appollo with one i860 displayed a radiosity/texture
mapped room at about 4 frames/sec. The 68030 called libraries on the i860
subsystem.
Finally, the price for such a board in the Amiga would be no less than (list)
$2500 in quantity. The 25 Mhz i860 (parts that can't handle 33 or 40 Mhz)
are still around $250. The RAMDAC will be over $100, VRAMS, dual port RAMS,
DRAMS, PALs, PLAs, etc, all add up. Also, if you want to design the board
well, you'll need a software tool like MENTOR (or equivalent, which require
a Sun, Appollo, or HP workstation to run). MENTOR goes for around $40,000.
To make your life easier, Intel offers a complete i860 design for the PC
(just a math coprocessor, no video).


John

[BTW, there is at least one company working on an i860 board for the Amiga]

Sean Cunningham

unread,
May 31, 1991, 12:16:15 AM5/31/91
to
In-Reply-To: message from cazabon@hercules


The closest thing to a "dream card" that I've heard of is the new one from
GVP, the PVA (hmmm...wasn't Commodore showing a product at trade shows
about a year ago called the PVA, with similar specs? Hmmm)

In addition to using a TMS34020 and providing flicker-free 24bit RGB in
overscan, with genlocking, it also provides for full-motion video in a
window on the WorkBench!!! It's supposed to come bundled with a 24bit
version of MacroPaint, and they're going for Caligari as well.

According to AmigaWorld, it'll be out this summer for about $2K. Not to
bad, and it aught to wax the Mac boards that boast similar specs.

Sean
/\
RealWorld: Sean Cunningham / \ "Doing our business is what
INET: se...@pro-party.cts.com VISION Amigas are for."
Voice: (512) 992-2810 \ /
// \/ "Holy #@*!" - any Psygnosis
KEEP THE COMPETITION UNDER \X/ GRAPHICS game player

LE...@qucdn.queensu.ca

unread,
May 31, 1991, 4:20:01 PM5/31/91
to
In article <1991May29.2...@nntp-server.caltech.edu>,

nyg...@nntp-server.caltech.edu (Michael T. Nygard) says:
>
>I've been looking at some numbers on the board. Here's what I came up with.
>
>Without using a CLUT (yuk.), assuming 80ns RAM, 1024x1024x24 resolution.
>
>1/(1024*1024*24*60) = .66227 ns/bit.
>(80 ns/cycle) / (.66227 ns/bit) = 120.8 bits/cycle
>
>That means a bus _at least_ 128 bits wide. A 128 bit bus results in 5% of
>the memory cycles availible for the processor.

Hmmmm. What about subdiving the 24 bit planes into smaller portions ? eg
1, 2 or 4 bit planes per bus or better still - per board. They'll require
a lot less bandwidth per bus so something like 32-bit or 64-bit should be
enough to allow for getting some CPU/graphics coprocessor cycles. Some sort
of multiplexor scheme would allow the CPU to talk to one of these buses at
a time. It would also be possible (not necessarily feasible) to allow the CPU
to access pixels on a word basis by accessing the same pixel location
on all the boards simultaneously. The multiplexor circuit is going to be
messy. :( Might be able to put it in a gate array. )

All of these bit planes would be combined to the D/A converter. (Of course
they are required to sync up to the same DOT clock for generating the pixels.)
The individual boards are identical so the user can upgrade to more bit-planes
if they need it or have $$$. It is very expandable in that aspect. On the
hardware side, each of these boards can have its own graphic processors to
speed things up. (Most bit-plane operations can be run in parallel...)

>
>Other problems:
>I hadn't planned on this ever being a commercial product. Getting the kind
>of printed circuit board this would require in single quantity would be very
>expensive.

My scheme allows you to build 1 bit plane and test it out. The boards are
much smaller and probably easier to layout. You would just have to order 24x
more boards when people start ordering from you.

>
>As for the cost estimate, it's hard to say. Could be real high, maybe not.
>There are too many factors at this stage (early design stage). No less than
>$1000 for a one-of-a-kind. Now if we could put together a few people...
>
>This probably belongs in .hardware, but it started here, and I think the
>main interest is here.
>
>Free free to contact me if you have suggestions (or would like to help!)
>
>--
>--------
>Mike Nygard
>nyg...@coil.cco.caltech.edu

K. C. Lee
"Be careful of what you wish for. You might get it."-Master Splinter,T.M.N.T.

Jay R. Reher

unread,
Jun 4, 1991, 2:43:06 PM6/4/91
to

Well such beasts do exist in the newer graphic workstations put out
by IBM, Silicon Graphics etc.. Silicon Graphics has one that is 1280 X 1024 X
24 bit, plus does 3 million gourad shades polygons per second and includes z-
buffering etc.. I am not sure of the bus size tho.. And by the way these bad
boys are not cheap. If you are wanting one for the amiga, have fun designing
it....


Jay R. Reher

Richard Blewitt

unread,
Jun 4, 1991, 4:34:30 PM6/4/91
to
In article <1991Jun4.1...@ucunix.san.uc.edu> re...@ucunix.san.uc.edu (Jay R. Reher) writes:
>
> Well such beasts do exist in the newer graphic workstations put out
>by IBM, Silicon Graphics etc.. Silicon Graphics has one that is 1280 X 1024 X
>24 bit, plus does 3 million gourad shades polygons per second and includes z-
>buffering etc.. I am not sure of the bus size tho.. And by the way these bad

Not quite, the top-of-the-line SG does ~1 million flat shaded
polygons (50 pixel tirangles) per sec.

>boys are not cheap. If you are wanting one for the amiga, have fun designing
>it....

_____________________________________________________________________

Not cheap? Doesn't everybody have $100,000 to buy a graphics board
for their own file servers (many K $$)? :)

Rick

The generic .sig Rick Blewitt rble...@ucsd.edu

Steve Brailsford

unread,
Jun 6, 1991, 1:03:53 PM6/6/91
to
nyg...@nntp-server.caltech.edu (Michael T. Nygard) writes:


>As for the cost estimate, it's hard to say. Could be real high, maybe not.
>There are too many factors at this stage (early design stage). No less than
>$1000 for a one-of-a-kind. Now if we could put together a few people...

$1000 for the board built? You got to be kidding. I worked on
a graphics board that didn't do nothing like what you are talking
about and it cost about $6000 to make one. This one would
cost at least $10,000 to make. Mass production is the only way
to make any money on these things. If you don't make 1000 or
more in one batch your wasting your time.
--
Steve Brailsford (f...@sparc.vitro.com) _____
Usenet: uupsi!vitro!sparc!fsb \/itro Corporation
Compu$erve: 73427,1466 14000 Georgia Ave.
Voice: (301) 231-1481 Silver Spring, MD 20906

Christopher Hassell

unread,
Jun 10, 1991, 8:19:21 PM6/10/91
to
nyg...@nntp-server.caltech.edu (Michael T. Nygard) writes:

[]Here's a description of my dream graphics board

[i860 processor for anti-aliasing lines, fast 3d polygon renders, Goroud & Phong
[] shading. (64 bit busses.)
[] 4M memory for frame buffer & z-buffer
[] 1024x1024x24 resolution
[] 8 bit z-buffer

[] Anybody know of such a beast, or anything similar. If not, I'll start doing
[] some designs.

No not really. I suppose it would be nice.

Basically as I think of it. There are alot better things that you could do
than to put a fast beast in alot of RAM and hook it up to a fast something else.

What would anyone say if you could split up the screen like CRAZY insted of
this constant blitting that is constantly used! Use a processor which can
constantly heave out the next address to memory for the screen to use (as long
as it is chip I suppose).

If you could map memory in arbitrarily complicated ways into the screen, you
could also do wierd wierd stuff like mapping many places all to one repeated
(binary-masked at least). Heck.. you could have live video without wasting a
single cycle (as long as it redirects the frame-buffer directly to the screen).

Memory is already used up with tons of graphics overlays and stuff.. blitting
is just a way to act like you are running sprites all over.

For another neat effect... map "colors" to sections of memory as well (maybe
even hardware ports?), and you can get "sparkling" backgrounds or shifting
backgrounds or lots of different effects, with only a little task out
somewhere shifting the bytes around if need be.

The screen IS like a high-high-speed serial device. It'd be nice if it
could be treated that way and used to its fullest extent.

In some ways... since a read-write op is rather common for memory chips, one
could blit AS sections of memory are being shown. (no screen-wipe that way!)

I cannot imagine anything better for graphics that doesn't boil down to
chewing through the numbers faster. (i.e. three-dee stuff is just that with
transformations and stuff simply going back into number-crunching).

[]Replies & suggestions are welcome!

I cannot say it wouldn't be nice to havce such a board supported. It'd be
downright fun if it could work and be a little bit CHEAPER than a large car.
.. but then that's what Amiga is for. ?:->

[]--
[]--------
[]Mike Nygard
[]nyg...@coil.cco.caltech.edu
--
If you are getting mail from me, my apologies for confusiong you ##############
### C>H> ### boulder.colorado.edu!tramp!hassell #### "Nietzche is dead." - God
--
If you are getting mail from me, my apologies for confusiong you ##############
### C>H> ### boulder.colorado.edu!tramp!hassell #### "Nietzche is dead." - God

Cybernetworx

unread,
Jun 15, 1991, 10:30:45 PM6/15/91
to
In article <hassell....@tramp.Colorado.EDU> has...@tramp.Colorado.EDU writes:
>nyg...@nntp-server.caltech.edu (Michael T. Nygard) writes:
>
>[]Here's a description of my dream graphics board
>
>[i860 processor for anti-aliasing lines, fast 3d polygon renders, Goroud & Phong
>[] shading. (64 bit busses.)
>[] 4M memory for frame buffer & z-buffer
>[] 1024x1024x24 resolution
>[] 8 bit z-buffer
>
>[] Anybody know of such a beast, or anything similar. If not, I'll start doing
>[] some designs.
>
>No not really. I suppose it would be nice.

Hmm, am I the only person to have heard of the Innosphere graphics board
for the amiga? If I recall properly, sometime about a year ago, a fellow
posted up a message describing just such a board that he had been designing.
I am positive I have the message buffered somewhere on my HD that gave
full specs on the board. I do remember that it came with options to have
either 1 or 2 i860's running in parallel. It supported 1280x1024x24 (?)
and 1024x1024x32 (?). It did say that it performed at roughly 70 Mflops
with 1 i860. Could render at 96 bits per pixel. Hmm, I'll have to dig
it up...but has anyone else heard about this, or remember the post? At
the time it was posted, I remember the guy said it wasn't completed but
was almost finished. He also said price would be between $5000-$10000.
I think the low end configuration came with 4 megs of ram, and also came
in 8 and 32 meg versions. This is all from memory, and this *WAS* a long
time ago, so there is a chance that some of my numbers/facts are wrong.

--
__
__ ///
\\\ /// UUCP: ...uunet!utai!mnetor!becker!cccan!entity
CYBERNETWORX INET: ent...@cccan.UUCP

Mark Thompson

unread,
Jun 17, 1991, 2:43:19 PM6/17/91
to
ent...@cccan.uucp (Cybernetworx) writes:
>>nyg...@nntp-server.caltech.edu (Michael T. Nygard) writes:
>>i860 processor for anti-aliasing lines, fast 3d polygon renders, Phong

>>[] Anybody know of such a beast, or anything similar.
>
>Hmm, am I the only person to have heard of the Innosphere graphics board
>for the amiga? If I recall properly, sometime about a year ago, a fellow
>posted up a message describing just such a board that he had been designing.

The project is defunct. The developer could not get the financial backing
to proceed with the product. I spoke with him a few weeks ago and he mentioned
that the whole thing had been handed off to the University of Lowell for
further tinkering with some involvement from Commodore. He didn't seem too
confident that anything would come of it soon. I find this most unfortunate
since an i860 based board is what I most need right now. The i860 is by far
the most bang for the buck out there and I am deeply disappointed that no
one has managed to bring one to the Amiga market, especially now with the low
cost 25MHz version ($175 quantity 1000). I design i860 based graphics systems
at my day job which pretty much prevents me from doing one for the Amiga,
but if anyone else is working on such a project, I would be more than happy
to quietly lend a hand on the development effort.
%~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~%
% ` ' Mark Thompson CONCURRENT COMPUTER %
% --==* RADIANT *==-- ma...@westford.ccur.com Principal Graphics %
% ' Image ` ...!uunet!masscomp!mark Hardware Architect %
% Productions (508)392-2480 (603)424-1829 & General Nuisance %
% %
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Jerry Thompson

unread,
Jun 24, 1991, 6:46:32 PM6/24/91
to

We have an i860 based board for the AT bus called the Horizon860. We have a
port of Pixar's Photo-Realistic Renderman for it. I have been trying to find
Amiga software which outputs RIB files. This would allow renderings to be
dumped to the Horizon860 for processing and then brought back to the Amiga
for display or displayed onto a TARGA+ or ATVista board on the PC side.

-Jerry
--
Jerry Thompson | // checks ___________ | "I'm into S&M,
I loved the peace and solitude | \\ // and | | | | Sarcasm and
so much, I invited my friends. | \X/ balances /_\ | /_\ | Mass Sarcasm."

Mark Thompson

unread,
Jun 26, 1991, 11:58:32 AM6/26/91
to
je...@truevision.com (Jerry Thompson) writes:

>ma...@calvin.westford.ccur.com (Mark Thompson) writes:
>>confident that anything would come of it soon. I find this most unfortunate
>>since an i860 based board is what I most need right now. The i860 is by far
>>the most bang for the buck out there and I am deeply disappointed that no
>>one has managed to bring one to the Amiga market, especially now with the low
>>cost 25MHz version ($175 quantity 1000).
>
>We have an i860 based board for the AT bus called the Horizon860. We have a
>port of Pixar's Photo-Realistic Renderman for it. I have been trying to find
>Amiga software which outputs RIB files. This would allow renderings to be
>dumped to the Horizon860 for processing and then brought back to the Amiga
>for display or displayed onto a TARGA+ or ATVista board on the PC side.

Yes, I am familiar with it. It was my understanding however that it cost
around $9000. I know that i860 based designs are not cheap but I was hoping
to see something a bit less lofty in price, something around $4K would be
reasonable. I know of at least one company that is OEMing i860 based graphics
boards for around $2K (but that is on a VME bus).

Reply all
Reply to author
Forward
0 new messages