Click Here
home features news forums classifieds faqs links search
6071 members 
Amiga Q&A /  Free for All /  Emulation /  Gaming / (Latest Posts)
Login

Nickname

Password

Lost Password?

Don't have an account yet?
Register now!

Support Amigaworld.net
Your support is needed and is appreciated as Amigaworld.net is primarily dependent upon the support of its users.
Donate

Menu
Main sections
» Home
» Features
» News
» Forums
» Classifieds
» Links
» Downloads
Extras
» OS4 Zone
» IRC Network
» AmigaWorld Radio
» Newsfeed
» Top Members
» Amiga Dealers
Information
» About Us
» FAQs
» Advertise
» Polls
» Terms of Service
» Search

IRC Channel
Server: irc.amigaworld.net
Ports: 1024,5555, 6665-6669
SSL port: 6697
Channel: #Amigaworld
Channel Policy and Guidelines

Who's Online
14 crawler(s) on-line.
 137 guest(s) on-line.
 2 member(s) on-line.


 matthey,  cdimauro

You are an anonymous user.
Register Now!
 cdimauro:  11 secs ago
 matthey:  2 mins ago
 agami:  7 mins ago
 Hypex:  12 mins ago
 Rob:  3 hrs 52 mins ago
 Dwyloc:  4 hrs 3 mins ago
 towo2099:  5 hrs 24 mins ago
 Marcian:  5 hrs 29 mins ago
 amigakit:  6 hrs 14 mins ago
 michalsc:  6 hrs 43 mins ago

/  Forum Index
   /  Classic Amiga Hardware
      /  Missed opportunities to improve the Amiga chipset – 6: the alternative of 16-bit innovations
Register To Post

Goto page ( Previous Page 1 | 2 | 3 )
PosterThread
cdimauro 
Re: Missed opportunities to improve the Amiga chipset – 6: the alternative of 16-bit innovations
Posted on 5-Sep-2024 5:35:18
#41 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3980
From: Germany

@bhabbott

Quote:

bhabbott wrote:
@Hammer

Quote:

Hammer wrote:

CDTV-CR's hardware C2P wasn't supposed to exist.

As far as we know it didn't.

Quote:
CDTV-CR has two extra custom chips not found in other Amigas i.e. Beauty 391246-01 and Grace 391245-02 (includes Gary/Gayle role)

CDTV-CR also includes a system on a chip CSG 4510 that was developed for C65.

According to the Amiga Hardware Database:-

"the newly introduced custom chips, Grace and Beauty, both have their equivalents in the original CDTV."

"Beauty drives the LCD display on the front panel, handles the buttons and the infrared port. It communicates with the rest of the system through the 4510 microcontroller chip."

"Custom chips
...
Grace - system address decoder
Beauty - front panel controller"

I doubt any C2P functionality is hidden in Beauty, given that it isn't connected to the system bus. Denise is exactly the same chip as in the A600. So that only leaves Grace, the 'system address decoder'. If, as you say, it had similar functionality to Gayle, it might have been possible to put a C2P converter it - though it would need a 16-bit data bus connection to be efficient (Gayle only has an 8-bit control port).

Good questions. Where's this logic located?
Quote:
@cdimauro

Quote:

cdimauro wrote:

The CDTV-CR comes out and lo and behold there is this bitplane-to-pixel converter

What's a 'bitplane-to-pixel converter'? What every Denise has in it, that's what! We need more than this terse description to show that the CDTV-CR had hardware C2P (that's chunky to planar, not 'bitplane to pixel').

Indeed. As it's written, it's the exact opposite: P2C.
Quote:
I also question what use it would have be been, given that even 6 bitplanes was a severe strain on performance. I wonder what they expected it to be used for.

Well, we used it for Fightin' Spirit...
Quote:
In any case, the C2P circuitry in Akiko must be significantly different from anything in the CDTV-CR. Even if Grace had something similar in it, they would still have to create an 8 bitplane 32-bit version that could be squeezed into Akiko.

No, the C2P can work with 16 bits at the time as well (hence, converting 8 16-bit packed values to 6 bitplanes).


@Hammer

Quote:

Hammer wrote:
OCS/ECS's "dual playfield" feature has up to 6 bitplanes.

That's TOTALLY IRRELEVANT, since those are TWO screens and not one!

How can you use the C2P logic there?!?

Bah...

 Status: Online!
Profile     Report this post  
cdimauro 
Re: Missed opportunities to improve the Amiga chipset – 6: the alternative of 16-bit innovations
Posted on 5-Sep-2024 5:40:38
#42 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3980
From: Germany

@Hammer

Quote:

Hammer wrote:
@cdimauro

Quote:
Neither AGA, with the exception of only fetching the content for display & sprites.


1. AGA is good enough to display 61 fps 320x200p 256 colors Quake demo1 with a fast enough CPU (e.g. CM4 Emu68 @ 1.8Ghz).

Using the same CPU power (RPi 4B Emu68 @ 1.8Ghz), OCS/ECS can do 28.9 fps EHB or 25 fps HAM6.

AGA can do 30 fps 640x200p 256 colors Quake demo1 with a fast enough CPU.

The Quake test shows AGA's 32-bit CPU to 32-bit Chip RAM link is 2X over the OCS/ECS counterpart.

For 2D beat-em-up, OpenBOR's AGA display is also tested with similar results.

It's faster than my recycled ET4000AX ISA on Pentium 166Mhz.

My point with Quake and OpenBOR tests, as long the composer processing source can generate frames at a certain pace, AGA is good enough until the release of Diablo's or StarCraft's 640x480p 256 colors which recommends a Pentium class CPU with PCI and DirectDraw accelerated SVGA card.

Where is the time machine so that people can use PiStorm with their Amigas?
Quote:
It wouldn't be to your liking since AGA's primary 2D composer hardware wasn't substantially updated i.e. AGA Blitter is about 60 percent faster due to memory bandwidth increase.

I've already written this to you: that's completely false! It's only 40%.

I've made PRECISE calculations here and reported the data. Why don't you get it, finally?!?

Since you continue to write the same crap all the times, can you PROVE it that it's 60%? I'm preparing other popcorns...
Quote:
Like ET6000's strong 2D performance, AAA's stronger 2D performance was canceled due to 3D considerations.

AGA would be a dead duck in 1996.


Quote:
2. AGA's target goal is entry-level SVGA's 640x480p with 256 colors from 16.77 million color palette since VGA's 640x480p 16 colors from 262,144 color palette were dated in the early 1990s.

In the early 1990s, many low-end SVGA cards have 16.77 million color palette is optional.

On paper, AGA looks reasonable.

Between ET4000AX vs C= AGA, I'm okay with AGA.

I'm not. AGA is crap, because you're beloved engineers made an HORRIBLE patch on top of the ECS, and used the 4 x bandwidth only for the display logic.

Only Amiga engineers make it possible!

 Status: Online!
Profile     Report this post  
cdimauro 
Re: Missed opportunities to improve the Amiga chipset – 6: the alternative of 16-bit innovations
Posted on 5-Sep-2024 5:46:45
#43 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3980
From: Germany

@agami

Quote:

agami wrote:
@cdimauro

Nice article with the appropriate amount of depth.

It would appear that other than external consultants missing from the equation, Commodore Amiga was a clash of Suits vs. Revolution Engineers. There doesn't seem to have been any or enough Evolution Engineers.

What Suits love are Evolution Engineers. They slowly inch forward a marketed product, providing just enough innovation while keeping the costs low and profit margins high. Unlike Revolution Engineers who have to be constantly reined in from their "Pie in the Sky" MO.

Revolution Engineers compromise via hacks. Evolution Engineers compromise by constantly mining and distilling ongoing R&D projects for useful and implementable innovations that fit timelines and budgets.

The right balance of the two types of engineers is critical, as Intel found out the hard way when having too many Evolution Engineers allowed AMD to revolutionise the desktop, HEDT, workstation, and server CPUs landscape.


The main problem with Commodore that it lacked engineers able to understand how the chipset worked and properly make it evolve with the technology progresses in a natural and coherent way.

Revolutions are certainly possible, but with the right vision and understanding. Not as an excuse because they weren't able to do their homework even for adding another 4 audio channels, for example, and propose a totally alien and super expensive DSP, instead.

TLDR; they were incompetent.

 Status: Online!
Profile     Report this post  
Hammer 
Re: Missed opportunities to improve the Amiga chipset – 6: the alternative of 16-bit innovations
Posted on 5-Sep-2024 8:22:01
#44 ]
Elite Member
Joined: 9-Mar-2003
Posts: 5846
From: Australia

@cdimauro

Quote:
That's TOTALLY IRRELEVANT, since those are TWO screens and not one!

How can you use the C2P logic there?!?

My point with CDTV-CR is to show how a rogue engineer is treated by Commodore.

Without the necessary management authority, it's useless.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

 Status: Offline
Profile     Report this post  
Hammer 
Re: Missed opportunities to improve the Amiga chipset – 6: the alternative of 16-bit innovations
Posted on 5-Sep-2024 8:35:09
#45 ]
Elite Member
Joined: 9-Mar-2003
Posts: 5846
From: Australia

@cdimauro

Quote:

@cdimauro
Where is the time machine so that people can use PiStorm with their Amigas?

Red herring. Focus on frame buffer transfers and display capabilities.

The complexity to compose the frame buffer can be reduced with different games.


Quote:

@cdimauro
I've already written this to you: that's completely false! It's only 40%.

Again, https://powerprograms.nl/amiga/cpu-blit-assist.html

32x32 bobs per PAL frame


A500 Blitter** = 11.

A1200 Blitter = 17 (improved with memory bandwidth upgrade).

17 / 11 = 1.54.

54 percent improvement.


----------

A1200 Blitter + CPU = 19

19 / 11 = 1.72

72 percent improvement.

I picked the modest 60 percent improvement.

Larger bob size can reduce the improvement gap between A500 and A1200. Your mileage can vary.


A1200 68EC020 = 8

Chip RAM gimped A1200 68EC020 CPU is 47% of A1200 Blitter.

Chip RAM gimped A500 68000 CPU is 72% of A1200 Blitter.

A no-brainer with CPUFastBilt patches with faster 68K CPU accelerators.


Quote:

@cdimauro
I'm not. AGA is crap, because you're beloved engineers made an HORRIBLE patch on top of the ECS, and used the 4 x bandwidth only for the display logic.

Wrong narrative. Without the necessary management authority, it's useless.

I'm bias towards texture mapped 3D games, not some dated 2D game like Fightin' Spirit.

Any method that improves 3D performance, I'm for it. Like many others, I don't care about 2D games during late 1992 to 1993.

For 1990s, Neo Geo sales flopped harder than 3DO's 2 million install base.

3DO has arcade quality Super Street Fighter 2 Turbo port. https://www.youtube.com/watch?v=vD5Guy7UKi8




Last edited by Hammer on 06-Sep-2024 at 02:28 AM.
Last edited by Hammer on 05-Sep-2024 at 09:20 AM.
Last edited by Hammer on 05-Sep-2024 at 09:05 AM.
Last edited by Hammer on 05-Sep-2024 at 09:02 AM.
Last edited by Hammer on 05-Sep-2024 at 08:54 AM.
Last edited by Hammer on 05-Sep-2024 at 08:49 AM.
Last edited by Hammer on 05-Sep-2024 at 08:47 AM.
Last edited by Hammer on 05-Sep-2024 at 08:46 AM.
Last edited by Hammer on 05-Sep-2024 at 08:43 AM.
Last edited by Hammer on 05-Sep-2024 at 08:39 AM.
Last edited by Hammer on 05-Sep-2024 at 08:36 AM.
Last edited by Hammer on 05-Sep-2024 at 08:36 AM.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

 Status: Offline
Profile     Report this post  
cdimauro 
Re: Missed opportunities to improve the Amiga chipset – 6: the alternative of 16-bit innovations
Posted on 6-Sep-2024 5:50:06
#46 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3980
From: Germany

@Hammer

Quote:

Hammer wrote:
@cdimauro

Quote:
That's TOTALLY IRRELEVANT, since those are TWO screens and not one!

How can you use the C2P logic there?!?

My point with CDTV-CR is to show how a rogue engineer is treated by Commodore.

Without the necessary management authority, it's useless.

And again this is ANOTHER, COMPLETELY DIFFERENT THING!

You're desperately trying to change the context again, talking of something completely different, to distract from the big load of b@alls that you've written before. RED HERRING!

And to be clear, since you live in the land of confusion, how the above engineer was treated or not had NOTHING TO DO with the TECHNICAL stupidity that you've reported before:

OCS/ECS's "dual playfield" feature has up to 6 bitplanes.

In fact, the C2P is NOT applying to this special display mode of the Amiga chipset.

Because it has NO SENSE! AT ALL!!!

Who's the idiot that wants to convert TWO packed/chunky 8-bit screens to TWO 8 colours / 3 bitplanes Amiga planar screen?

Care to give an answer, dear desperate citizen of The Land of Confusion?

The reality is always the same with you: you've NO clue of the technical things of what you talk about.

You only write because you've a keyboard...

 Status: Online!
Profile     Report this post  
cdimauro 
Re: Missed opportunities to improve the Amiga chipset – 6: the alternative of 16-bit innovations
Posted on 6-Sep-2024 6:09:03
#47 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3980
From: Germany

@Hammer

Quote:

Hammer wrote:
@cdimauro

Quote:

@cdimauro
Where is the time machine so that people can use PiStorm with their Amigas?

Red herring. Focus on frame buffer transfers and display capabilities.

The complexity to compose the frame buffer can be reduced with different games.

Again, a complete non-sense. The frame buffer is NOT relevant when talking about the machines OF THE TIME.

We had what we had. AT THE TIME. Which means the chipset / framebuffer AND the CPUs. OF THE TIME.

It's totally irrelevant and a complete non-sense using a time machine to replace only the CPU of an old system with a MODERN one with steroids, to prove that the frame buffer is enough.

Do you know that even the opposite is possible? We can replace not even the frame buffer, but the complete graphic subsystem with a MODERN one to show that even a 68000 was able to show his muscles on doing games with Ray Tracing...
Quote:
Quote:

@cdimauro
I've already written this to you: that's completely false! It's only 40%.

Again, https://powerprograms.nl/amiga/cpu-blit-assist.html

32x32 bobs per PAL frame


A500 Blitter** = 11.

A1200 Blitter = 17 (improved with memory bandwidth upgrade).

17 / 11 = 1.54.

54 percent improvement.

Now you're comparing apples with oranges, In fact, the discussion was the Blitter. ONLY the Blitter.

Which that you've to remove all other variables, if you want to make a real comparison.

Specifically, you're comparing the Blitter running on an Amiga 500 = 7Mhz 68000 and 7Mhz 16-bit Chip RAM, with the same but running on an Amiga 1200 = 14Mhz 68EC020 CPU and 14Mhz 32-bit Chip RAM.

Obviously the test is showing the advantages coming ALSO from the different CPU, bus size and frequencies.

See below to get how tests should be made.
Quote:
----------

A1200 Blitter + CPU = 19

19 / 11 = 1.72

72 percent improvement.

Completely irrelevant: the topic was ONLY about the Blitter and NOT about adding the CPU.
Quote:
I picked the modest 60 percent improvement.

Mixing results (already biased: see above): Blitter and Blitter + CPU. Whereas the discussion was about ONLY the Blitter.
Quote:
Larger bob size can reduce the improvement gap between A500 and A1200. Your mileage can vary.

Guess what: the bigger the bobs, the lesser is the impact of everything else you're approaching the pure differences in how the Blitter works.

Which means: this is the way to measure the impact of the AGA 4x bandwidth compared to the OCS/ECS one, for the Blitter.
Quote:
A1200 68EC020 = 8

Chip RAM gimped A1200 68EC020 CPU is 47% of A1200 Blitter.

Chip RAM gimped A500 68000 CPU is 72% of A1200 Blitter.

A no-brainer with CPUFastBilt patches with faster 68K CPU accelerators.

Again, totally irrelevant.

BTW, you're so "intellectually honest" that you've edited your post and removed the results with the 64x64 BOBs.
Guess what: they were proving my thesis, since they have shown a 25% improvement of the AGA over the OCS/ECS

Poor man: you're desperately trying to hide the reality. Not a chance with me, looser.

Anyway, here are the real numbers: https://amigaworld.net/modules/newbb/viewtopic.php?topic_id=44362&forum=2&start=240&viewmode=flat&order=0#855920

Enjoy! And try to learn something in your life, ignorant!
Quote:

Quote:

@cdimauro
I'm not. AGA is crap, because you're beloved engineers made an HORRIBLE patch on top of the ECS, and used the 4 x bandwidth only for the display logic.

Wrong narrative. Without the necessary management authority, it's useless.

You need also capabilities, which the engineers haven't show...
Quote:
I'm bias towards texture mapped 3D games, not some dated 2D game like Fightin' Spirit.

Guess what: Amiga games were almost all about 2D games.
Quote:
Any method that improves 3D performance, I'm for it. Like many others, I don't care about 2D games during late 1992 to 1993.

And what's the problem? You were a rare bird on the Amiga landa, since it was almost all about 2D games.

You were using the wrong platform...
Quote:
For 1990s, Neo Geo sales flopped harder than 3DO's 2 million install base.

LOL. You don't know of what you talked about!

A flop?

It DOMINATED the arcade market. And sold a lot in the console market, taking into account the expensive hardware.
Quote:
3DO has arcade quality Super Street Fighter 2 Turbo port. https://www.youtube.com/watch?v=vD5Guy7UKi8

And... who cares?!?

BTW, it was a 2D game, eh! Whereas you clearly statedf that you only liked 3D games in that period. "Coherent"...

 Status: Online!
Profile     Report this post  
Goto page ( Previous Page 1 | 2 | 3 )

[ home ][ about us ][ privacy ] [ forums ][ classifieds ] [ links ][ news archive ] [ link to us ][ user account ]
Copyright (C) 2000 - 2019 Amigaworld.net.
Amigaworld.net was originally founded by David Doyle