@thread

IMHO whilst the CD32 was an interesting if limited machine, a far more interesting scenario would have been if Commodore had been involved the 3DO project.

After all given the history behind the men who designed the 3DO, this felt more like a proper next-gen Amiga, rather than an A1200 with a CD drive tagged on.

At the time, a Commodore released 3DO machine at around £300 would have flown off the shelves.

_________________
Amigaone X5000/20 4GB Radeon RX 550 Polaris 12 AmigaOS4.1 Final Edition Update 1
Amiga 1200 Workbench 3.1.4
Amiga CD32

@outrun1978

How would C= have knocked £100 of the price with their involvement? Their CSG/MOS semiconductor chip fabs were out of date and their engineers were few in number and involved in patching up/cost reducing AGA Amiga tech rather than pushing the bleeding edge like the 3DO was trying to do!

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

agami Quote:

The one I voted for is "Underpowered". Not only because of the lackluster A1200-based spec + Akiko, but because it was the closest thing to not having the overall hardware platform required to properly enter into the console market. ... The console would quite possibly have a minimum of 030, and an Akiko that wasn't half baked. The only expansion that belongs in the living room is the FMV module. And they should've kept the PCMCIA for some future-proofing. E.g. Adding a modem for dial-up multiplayer. And possibly reintroduce the Joyboard for some unique living room fun for the whole family.

The 1993 CD32 had better hardware than the SNES (1991) or Mega/SegaCD (1991) but was inferior to the 3DO (1993), Atari Jaguar (1993), Sega Saturn (1994), Sony Playstation (1994) and Nintendo 64 (1996). These consoles had relatively weak and low power CPUs compared to x86 desktop brute force gaming hardware driven by Doom and Quake. A more powerful CD32 CPU could have been used to offset the aging and lackluster AGA chipset hardware until later AA+ and 3D enhancements. A 68030@28MHz would have been more powerful than most of the console CPUs moving the CD32 from near the bottom in CPU performance to near the top but would have added to the cost.

Console | CPU@MHz | Caches in Bytes
SNES 65816@3.58MHz -
SegaCD 68000@12.5MHz -
CD32 68EC020@14MHz 256I
3DO ARM60@12.5MHz -
Atari Jaguar Custom RISC@26.59MHz
CD32 68030@28MHz 256I+256D
Sega Saturn Hitachi SH-2@28.6 MHz 4096U
PS1 MIPS R3000A@33.87MHz 4096I+1024D

The 68020 was released in 1984 and was older than most of the newer RISC CPUs. It was easier to program and reduced memory traffic with good code density and the 256B instruction cache but it was weaker per clock than the often deeper pipelined RISC CPUs so needed to be clocked up for performance. The 68030 is a similar design to the 68020 but with the following advantages.

o higher clock speeds up to 50MHz
o ~5% better performance and 25% lower power at the same clock speed compared to 68020
o 256B data cache with burst mode to caches (further reduces memory traffic)
o synchronous bus can accelerate memory performance by up to 33%
o MMU (not on 68EC030)

A 68030@28MHz could more than double CPU performance compared to a 68020@14MHz and would likely fit the CD32 power budget. Let's look at cost feasibility from 1992 Dataquest projected data for average sale prices (ASP).

68020/68EC020
1990 $80
1991 $55
1992 $37
1993 $24
1994 $15
1995 $11
1996 $9

68030/68EC030
1990 $125
1991 $95
1992 $55
1993 $39
1994 $31
1995 $28
1996 $25

68040/68EC040/68LC040
1990 $450
1991 $390
1992 $260
1993 $169
1994 $118
1995 $95
1996 $90

68060/68EC040/68LC040
1994 $350 (new PPC 603 was projected to sell for $100 in 1994; insane profit margin difference?)
1995 $245
1996 $184

A 68030 was likely only about $15 more than a 68020 in 1993. A 68EC020@14MHz is a low clocked 68020 and a 68030@28MHz would be a little above middle for clock ratings of 68030s likely adding to the difference. A surface mount 68EC030 is probably the best candidate but Motorola only has 25MHz and 40MHz rated CPUs where the full 68030 includes 33MHz rated CPUs. It looks bad to overclock a 68EC030 to 28.64MHz (NTSC) which is 3.64MHz or 15%. A special arrangement with Motorola could have been worked out or another oscillator and a little more logic added to clock the CPU separate from the chipset. Total cost for a CD32 68030 CPU with more than double the performance is perhaps $20-$30 more.

Console CPUs had small or no caches making them tricky to program and resulting in the CPU contending with the custom chips for memory bandwidth. Sometimes separate memory was used but this wasted expensive memory. Sometimes higher bandwidth/performance memory was used like VRAM, EDO or SGRAM but this added cost. The CD32 used the cheapest possible memory and was at the low end of memory quantity with 2MiB for the new generation of consoles released starting in 1993. It suffered from sharing memory bandwidth between the CPU and custom chips which was made worse by cheap slow graphics memory. The CPU has to wait on the custom chips unless fast memory is used which Amiga users know boosts performance considerably. A 68030 CPU upgrade for the CD32 would lose some of its performance boost without fast memory. One MiB of cheap memory in the early 1990s could be purchase for as cheap as $30 and C= likely could buy it in quantity for cheaper. Perhaps C= could have added the 68030 and 1MiB of fast memory for $50-$75 their cost that would have doubled to tripled performance. The performance would have been much more impressive and allowed many newer games coming out for other consoles and computers but it would have been more difficult to compete with consoles with better chipsets as the cost climbed closer. Worth it?

C= likely realized the CD32 would be barely competitive in performance and features for a console. The Akiko chip chunky kludge instead of AA+ chipset chunky support shows they had half a clue they were behind the technology curve. A 68030@28MHz would have made the CD32 less of a dog but maybe they didn't want to endanger Amiga 1200 sales which would have then looked inferior unless the same CPU upgrade was given to them too. Ideally, the Amiga 1200 would have been introduced with a 68030@28MHz to be more competitive with x86 hardware but C= was already in financial trouble, perhaps the reason the CD32 hardware was lower spec. Adding 1MiB of fast memory to the CD32 may have been a cheaper way to boost CPU performance while leaving more chip memory, chip memory bandwidth and overall memory for games. Cost was important and I believe C= planned on reducing the hardware cost lower than more powerful competitors while leveraging the unique advantages of the CD32 and large Amiga games library that could be converted. The Amiga 500 cost $200 to produce years earlier which the simpler CD32 could likely reduce.

$60 2MiB of chip memory
$30 68EC020
$15 CD-ROM
? AGA chipset and support chips
? motherboard
? case and controller
? power supply
? assembly cost

I expect the CD32 would have remained profitable in the $150-$200 U.S. range. I bought my CD32 with Chaos Engine and Microcosm for $150 in the U.S. shortly after the C= bankruptcy was announced but that was likely a "reduce to clear" sale. With cheap games and expansions, it offers pretty good value in that range. If may have even been possible to reduce the cost enough to sell in the $100-$150 range increasing survival chances as a budget console alternative but that would have been difficult with a 68030 and 1MiB of fast memory upgrade and the introductory retail price may have been $449 instead of $399. This may have been acceptable against the $699 3DO but doesn't look so hot against the $299 PS1 which arrived in 1995 for most of the world. If C= had an upgraded CD32 with 68EC030 and AA+ chipset by 1995, they would have the option of licensing the 68060 and 3dfx Voodoo chipset for 3D which would destroy the PS1 spec. The big question is how cheap the 68060 could be provided after the AIM agreement. It looks to me like Motorola was taking a reduced profit margin on low end PPC CPUs to subsidize PPC adoption which may have backfired when the PPC 603 cache performance required doubling the cache sizes and a die shrink into the PPC 603e but it still probably cost less than the EOL 68060 Motorola was milking. CPU performance can go a long way in making up for hardware deficiencies as Intel realized with x86 on the desktop, but it was too hot to put in a console at that point. It is also easier to incrementally upgrade CPU performance retaining compatibility as the new generations of x86-64 consoles are doing. I'm not saying C= was planning similar in the 1990s but they should have been.

Leaving the PCMCIA slot on the CD32 would have been awesome from a user perspective as a flash card could have been used like a hard drive and would have allowed very fast booting. This would have provided a lot more space for saves than the small non-volatile memory too.

agami Quote:

Also, you can't just drop console hardware on the market like it is just another personal computer. The console would need to sell at or near cost, and the profits would come from game publishing licenses: Like every other game console maker. That way, it would look more appealing and not much more expensive than an SNES or Sega Mega Drive (Genesis). And then by 1995 it could've still been sold by being priced lower than the Sega Saturn and the Sony PlayStation.

The CD32 was much easier and cheaper to develop for than other consoles. Let's compare the CD32 with 3DO and PS1.

CD32
Dev system: ~$3,000 (Amiga 4000/040 with MMU but could be an Amiga 1200 with accelerator)
Software Royalty: $3/disc

3DO
Dev system: ~$10,000 (officially Mac but early and often Amigas)
Software Royalty: $3/disc
Hardware Royalty: $?/certification

PS1
Dev system: ~$18,000 or ~$50,000 (conflicting sources)
Software Royalty: $10/disc

CD32 Developer Notes
http://www.kaiiv.de/servicemanuals/CD32%20Developer%20Notes.pdf
https://www.linkedin.com/pulse/interview-reichart-von-wolfsheild-ashley-wing

Per game royalty advantages and disadvantages
+ cheaper hardware
+ encourages higher quality games
- fewer developers and games
- more expensive games
- closed hardware often chosen to deter non-royalty paying software and uses of subsidized hardware

I see the benefits of open hardware offsetting the advantages of cheaper subsidized hardware. The CD32 had a small $3/disc U.S. royalty fee but the hardware didn't seem to depend on a subsidy because it was already cheap enough so C= practically encouraged embedded use and expansion as a value added benefit. The CD32 could be turned into and used as a full general purpose computer and, despite the short life, it had already attracted several embedded customers despite the much smaller embedded market then. A large portion of Raspberry Pi sales are into the embedded market because it is open and I believe this was a strategy from the beginning to boost volume sales for mass production even though they don't have a gaming platform like the Amiga. I would rather have more developers and more and cheaper games than higher quality games too. Reviews are easier to find today and could be incorporated into an "official" online store. The Amiga lack of security and openness makes the open hardware option easier as well.

Some people blamed the 3DO failure on too high of hardware cost because of not enough hardware subsidization with the low $3/disc royalty. The low fee was attractive to game developers which 3DO needed being a new business with no reputation or brand name in the console market. 3DO tried to make money off licensing the hardware to manufacturers and charging a royalty for each console sold but I believe this is partially responsible for the high hardware cost. The hardware royalties and profit margin added by the manufacturers likely partially offset the competitive cost savings found by different manufacturers. Maybe there was just too much hardware driving up the cost even though it was cheaper than a x86 desktop computer with similar capabilities (x86 desktop hardware was more versatile, open and expandable while 3DO was a dedicated games machine). CDTV was a failure at a high price too but the better priced CD32 would have been a success had C= not failed.

agami Quote:

Commodore had the right, if not perfect engineering team to be competitive in the gaming console market. They just happened to also have zero business acumen required to properly enter this market. And the company being in dire straits is no excuse for doing a half-arsed job.

The CD32 console market launch could have been more professional with better planning of Amiga development from the moment C= acquired the Amiga but it was still successful on a very low budget as C= approached bankruptcy. Had C= survived, the CD32 may have survived longer than higher performance and superior hardware alternatives like 3DO and the Sega Saturn. Yes, C= had the "engineering team" to be competitive but falling behind by years in technology makes it difficult to catch back up even with good leadership. The CD32 demonstrates the advantage of a simple, efficient, expandable, low cost and compatible product overcoming trailing edge technology hurdles though.

Last edited by matthey on 10-Sep-2023 at 02:19 AM.

@matthey

Putting a fast 030 or even 040 would have been a total waste without adding at least some Fast RAM which would have required some extra logic and more (smaller) RAM chips to have it 32Bit width.

Not feasible without a price increase, and since most CD32 games were just repackaged A1200 games few would have used the extra power.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@BigD

How do you get dpaint anims out of the THEA500?

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

Kronos Quote:

Putting a fast 030 or even 040 would have been a total waste without adding at least some Fast RAM which would have required some extra logic and more (smaller) RAM chips to have it 32Bit width.

I agree. Wasn't that one of my points above? Adding 1MiB of fast memory isn't much cheaper than upgrading to a 68EC030 with more than twice the performance but may be better because it reduces the time the CPU waits on the chipset accessing chip memory, more free chip memory is available for chipset data and 50% more total memory would have allowed some newer ported games. If adding a higher performance CPU, 1MiB of fast memory with it gives performance synergies and it is a partial waste without it.

A 68040 would have cost too much likely adding $150 or more to the cost of the CD32. The 68040 uses too much power and runs hot except for the 68LC040 which drops from 5V to 3.3V and eliminates the FPU. Eliminating the FPU wasn't a problem as consoles used mostly fixed point integer even up to the PS1. This was fast and allowed 3D to be cheap and low power enough but was also poor quality. The x86 desktop brute force fps games started using the FPU and had better quality 3D games but needed clocked up P5 Pentiums. Desktop 3D cards started coming out about a year after the PS1 with the November 6, 1995 3dfx Voodoo card which had superior 3D quality compared to the PS1. Perspective correction stopped the lens warped looking images and subpixel correction stopped the bouncing edges of objects. The 3dfx Voodoo chipset may have run hotter compared to the PS1 chipset but that can be solved with a chip die shrink or two. In fact, Sega considered using the 3dfx Voodoo chipset for the Dreamcast.

https://en.wikipedia.org/wiki/3dfx#Dreamcast Quote:

In 1997, 3dfx was working with entertainment company Sega to develop a new video game console hardware platform. Sega solicited two competing designs: a unit code-named "Katana", developed in Japan using NEC and Imagination Technologies (then VideoLogic) technology, and "Blackbelt", a system designed in the United States using 3dfx technology.

That could have been C= had they survived as the Amiga had the 2D chipset and 3dfx the 3D chipset that it was trying to license.

Kronos Quote:

Not feasible without a price increase, and since most CD32 games were just repackaged A1200 games few would have used the extra power.

Right. C= was primarily counting on leveraging the existing Amiga game library with some enhancements for the huge CD storage space. These games used the Amiga philosophy of offloading the CPU and the C= philosophy of using the cheapest possible hardware. The chunky support in Akiko is not needed for this though. This is a clue that C= hoped they would get some new and ported games despite the low CD32 spec compared to the competition. They did get some ports although some may have been planned AGA ports for the Amiga 1200 already in development. Some of them looked good but were borderline too slow like Wing Commander and UFO: Enemy Unknown which would have been more playable and enjoyable with a little more performance. There is no free lunch though with a console where PPA balancing and cost concerns are more like embedded hardware than desktop hardware.

@matthey

Quote:

matthey wrote: Right. C= was primarily counting on leveraging the existing Amiga game library with some enhancements for the huge CD storage space.

I don't think it was that complicated.

By 1992 C= had some serious cash flow problems and couldn't even build enough A1200 to satisfy the relatively low demand.

This did lead to even less cash in 93 and the rest is history.

So when they launched the CD32 they had already missed the window on getting the essential launch titles and lacked the funds for both making the HW in numbers or removing that BS-patent roadblock to sell in the US.

-> just throwing something on the wall hoping it might stick. It didn't and we got the Deathbed Vigil video.....

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@kolla

Quote:

@BigD How do you get dpaint anims out of the THEA500?

You use AMiNIMiga, save your anim to the USB. Move the files to the Shared folder within AMiNIMiga. Then copy the anims off the USB on another device (the shared folder is visible on other devices). Simples!

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@matthey

It certainly was an accomplishment to get a game console out in such a short space of time and at such a low cost of production. Reverse engineering an A1200 certainly helped.
Furthermore, that way C= didn't have exorbitant R&D costs associated with the unit, so the cost of a fully produced unit would've been less than $150.

Adding the faster CPU and some Fast RAM + license fee to Cad Track, would've resulted in the console's production price still less than $200.
So in a market with a SNES at $139, launching a console whose games don't look all that much better than SNES for $399, is commercial/retail suicide.
Without the extra hardware (underpowered), it was not going to encourage the development of better games, e.g. PS1.

Sure, existing Amiga gamers on older OCS systems would see it as a cheeper way to get into AGA gaming, but at almost 3x the price of a SNES, it's not going to sway anyone away from the dominant 16-bit systems.

The base system could've launched at $229 bundled with a couple quick A1200 AGA ports, and games could've still beat SNES title prices at a median $49, which would've allowed for a more profit friendly $6-9/CD royalty for Commodore.

At $229, there would've been an early install base at about 3x compared to $399, of a much better system (030/25 + 1-2MB Fast RAM) which would be an appealing target for NEW and potentially exclusive game development.

If awards were given for innovative hardware production on a shoe-string budget, then 1993's Commodore would win that accolade. As it is, what they produced amounted to no more than a proverbial fart in the wind.

_________________
All the way, with 68k

@BigD

Does it involve turning off the emulator?

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

You have to physically transfer the USB to your/PC/Mac/USB equipped Amiga since no one has hacked networking support yet! However, the 'shared' folder is visible under the AMiNIMiga if this is what you mean!?

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@kolla

NEWS FLASH! Internet enabled on THEA500 Mini!

Quote:

AMiNIMiga and Pandory team brings you Internet on your A500Mini 🙂 (And Amiberry 5! Yeah you heard right 🙂 ) Here is Pandory Teams AMiNIMiga Teaser: https://youtu.be/Ri7JKFFkm_U?si=C2oqvGUPp7bL86uf An amazing feat in bringing networking to the Mini! AMiNIMiga v210 Update will give you the chance to surf the web like its 1995! Stream Music Modules while chatting with other enthusiasts. Connect your A500Mini to your local PC for easy file transfers to and from AMiNIMiga. Chat with friends and foes over IRC with AmIRC. Amongst MANY other useful and fun things. Here is a incomplete changelog for AMiNIMiga v210, and this will grow more for the release, which will happen in the coming weeks.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@agami
SNES is an anomaly. The hardware was more behind the tech curve when launched in late 1990 than the CD32 when launched in 1993. It has a good console chipset but it is handicapped by the 16 bit 6502 family CPU and low memory. The SNES arrived in North America in 1991 and Europe in 1992 where the competition was appearing with CD-ROM drives. So was it game over for the SNES? No. It competed in the budget console market with very cheap hardware, sold over 49 million units becoming likely the most successful console ever (PS2 later surpassed) and the 1995 PS1@$299 market disruption did not keep it from surviving until 2005. Nintendo knows marketing and game play areas where C= was oblivious. The SNES and CD32 were both competing in the budget console market with very different strategies and hardware. There is more room in the budget console market as many more people can afford the hardware. Over 30 million units of the Sega Genesis were sold into this partially overlapping market. Surprisingly especially with retro gaming popularity, the huge budget console market is non-existent today although partially replaced by one and done Mini retro emulation eye candy from time to time, emulation on all kinds of computer devices and FPGA hardware for hardcore retro enthusiasts. It is possible to make the hardware cheaper than ever as the Raspberry Pi hardware has demonstrated and grow a "hobby" market to similar volumes as the SNES even with minimal retro gaming appeal. RPi sales may be more due to embedded markets than hobby and educational markets but that is where the CD32 was going along with providing a low cost general purpose hobby computer like the RPi. How successful CD32 and later iterations could have been depends heavily on integration and cost reductions but the CD32 is more powerful and much more versatile than the console specific SNES which has minimal upgrade potential. The SNES had more cost reduction potential in the early 1990s but technology was quickly closing the gap. Today, add a higher performance 68k CPU with more memory and a CD32 seems almost modern or at least RPi like. Doing the same to a SNES would just ruin the one asset it has, compatibility.

I agree that it would have been nice if C= had released the CD32 at a lower cost to grab more market share. The SNES was launched at $199 in 1990 while the CD32 was launched at $399 in 1993. The Japanese console businesses often liked to sell the hardware near or below cost to grab market share and then make it back later off game royalties. This is risky as more hardware inventory has to be ready at launch and production and supply chains must be ready to churn out more units. The financial situation at C= meant this option was less appealing. They chose a high introductory price with higher profit margins knowing they could quickly drop the price based on demand. THEA500 Mini was launched with a similar marketing strategy which maximizes profit for a quick to market one and done product where there is minimal competition. Had C= launched the CD32 at $349 or $299, it would have seemed less exploitive and the same can be said for the A500 Mini launching at a lower price. C= could have upped the spec to increase the value and still potentially launched at $349 but they likely already had 68EC020 CPUs in stock saving manufacturing costs. C= needed cash ASAP and their strategy sacrificed market share and cash flow later for a quick return much like THEA500 Mini. It's the opposite strategy of trying to gain market share and increase a growing user base. I'm still not a fan of subsidizing the hardware with game royalties. It may work for a pure game system if willing to take higher risks but it is deceptive with the decreased up front cost of hardware offset by the increased cost of games later.

Last edited by matthey on 11-Sep-2023 at 06:31 PM.
Last edited by matthey on 11-Sep-2023 at 06:29 PM.

@matthey

THEA500 Mini is now sub-£100, Internet enabled and already is a year plus ahead of the A600GS competition! I can't see that they've done anything wrong other than not courting Amiga resellers!

The SNES needed expensive chips added to cartridges to impress with StarFox. The CD32 would have done great with a 28MHz 030 and fast ram for £399.

Last edited by BigD on 11-Sep-2023 at 06:37 PM.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@BigD
The Retro Games Limited guys are marketing pros, THEA500 Mini continues to offer competitive value even if some of it comes from eye candy and they seem to understand and respect the retro Amiga even though the hardware could have been better like they wanted. Some things are too hard and beyond the control of mortals which seems to happen on the Amiga often.

CD32 > SNES
C= < Nintendo

@matthey

...Nintendo < their hype

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

BigD Quote:

...Nintendo < their hype

Nintendo was doing something right to sell inferior hardware. For a laugh, look at some of the most impressive SNES games using 3D hardware features but including non-console like games which the Amiga is good at.

The 15 Most Graphically Impressive 3D Super Nintendo (SNES)
https://www.youtube.com/watch?v=g3TydpBzUy0

I couldn't find a similar video for the Amiga but the best CD32 games include some non-console type of games. Most of the games have few enhancements compared to AGA or ECS versions.

The 30 Best Commodore Amiga CD32 Games Ever Made!
https://youtu.be/OT4MRPrAEow?t=61

Here are some comparisons of various consoles where the CD32 holds up against better competition than the SNES. SNES handles these games relatively well considering the low spec but is outperformed by the CD32 and older Sega Genesis.

Cannon Fodder -Versions Comparison- Amiga, CD32, AtariST, MS-DOS, Sega Genesis, and much more!
https://www.youtube.com/watch?v=fIUWEDw6sDk

Chaos Engine Version Comparison | Amiga - Atari ST - CD32 - MS-DOS - Mega Drive - SNES - Steam
https://www.youtube.com/watch?v=WtcgRm6z8ug

The CD32 wins easily for versatility but the SNES is good at sprite and tile traditional console games like Dennis the Menace which is better on the SNES. SNES tended to get higher quality game conversions. Sometimes the SNES would surprise with certain types of games.

Top Gear 2(Super Nintendo VS Amiga CD32)Side By Side Comparison Graphics
https://youtu.be/9NGml0OqXmc?t=132

I wonder how much more development time was spent on SNES conversions compared to CD32 versions. More CPU performance and memory is easier to use than chipset features and more memory makes chipset features easier to use as well. A good example is the SNES 8 voice 8 bit stereo sound with 16 bit DSP but only 64kiB of high performance SRAM compared to the Amiga 4 voice 8 bit stereo sound with DMA to play much larger samples in 2MiB of slow DRAM chip memory. The Amiga sounds so much better compared to the SNES and that is not even counting the CD32 16 bit CD-ROM audio that could be mixed in. The CD32 is underappreciated because it was short lived and could have been better but it is still impressive when receiving quality ports and holds up reasonably well even to higher performance but often more difficult to program competitors.

Last edited by matthey on 12-Sep-2023 at 02:46 AM.

@matthey

Quote:

matthey wrote: C= likely realized the CD32 would be barely competitive in performance and features for a console. The Akiko chip chunky kludge instead of AA+ chipset chunky support shows they had half a clue they were behind the technology curve. A 68030@28MHz would have made the CD32 less of a dog but maybe they didn't want to endanger Amiga 1200 sales which would have then looked inferior unless the same CPU upgrade was given to them too.

The A1200 could easily be upgraded with more RAM and faster CPU, and many owners did. I think Commodore was expecting this. I think they also expected that many wouldn't until they had to, just like many A500 owners didn't get a trapdoor RAM expansion until they had to.

The advantages of not making the CD32 more powerful than the A1200 include lower cost (essential) and a more homogeneous platform. But I'm betting 99% of it was cost. Why put expensive extra hardware in it when developers had barely begun to get the most out of a stock A1200?

Quote:

Ideally, the Amiga 1200 would have been introduced with a 68030@28MHz to be more competitive with x86 hardware

X86 hardware was a target moving so fast that there was no hope of keeping up with it for long. Better to put a bare minimum in the A1200, and let users upgrade as as faster hardware came out.

Realistically the only PC an A1200 could hope to compete against was a 16MHz 386SX. This was a popular low-end system in 1992, but around this time developers realized they could make enough sales by only developing games for the high-end (or even beyond), and owners of low-end systems would just have to buy a new PC if they wanted to play them. That started a trend which continues today. But that was not what Amiga and console owners were expecting. Even today we get pissed off when someone releases a new game that requires an 80MHz 060 or a Vampire, for no obvious reason than they were too lazy to make it work on a lower spec machine.

Quote:

but C= was already in financial trouble, perhaps the reason the CD32 hardware was lower spec.

It was designed to compete against consoles that were similarly priced or cheaper. Yes, Commodore was in trouble financially, but even if they weren't a higher price would be dicey. The biggest complaint against the 3DO was the price, and it was backed by companies with far more resources to burn than Commodore ever had.

Quote:

Cost was important and I believe C= planned on reducing the hardware cost lower than more powerful competitors while leveraging the unique advantages of the CD32 and large Amiga games library that could be converted. The Amiga 500 cost $200 to produce years earlier which the simpler CD32 could likely reduce.

I agree with that.

Quote:

Leaving the PCMCIA slot on the CD32 would have been awesome from a user perspective as a flash card could have been used like a hard drive and would have allowed very fast booting. This would have provided a lot more space for saves than the small non-volatile memory too.

Sure would, but was overkill in a machine that had to be pared down to the lowest possible cost. The CDTV had a memory slot that nobody used because the cards were way too expensive, and PCMCIA NV RAM cards were too.

My choice would have been a floppy drive interface, which would be more versatile and could be made quite cheaply. I designed one using a few standard TTL gates. Disk Drive Interface for CD32

Quote:

The CD32 was much easier and cheaper to develop for than other consoles. Let's compare the CD32 with 3DO and PS1. CD32 Dev system: ~$3,000 (Amiga 4000/040 with MMU but could be an Amiga 1200 with accelerator) Software Royalty: $3/disc

Most Amiga developers - and many amateurs who might want to become developers - already had a system powerful enough to do the job. So for many of us the dev system cost was $0. And since we already had vast experience on the Amiga the learning curve was essentially zero too.

Quote:

I see the benefits of open hardware offsetting the advantages of cheaper subsidized hardware.

I agree with that too. If only Commodore hadn't gone bankrupt and killed interest in the Amiga...

Quote:

Some people blamed the 3DO failure on too high of hardware cost because of not enough hardware subsidization with the low $3/disc royalty.

Consoles were like inkjet printers, the hardware was sold at cost and profits made from the (expensive) cartridges. That model didn't work so well with CD-ROM. Everyone knew how much it cost to press discs (less than typical royalties) and the hardware was bound to be more expensive because it had a CD drive (unless you were Sony, who made their own drives).

Quote:

The low fee was attractive to game developers which 3DO needed being a new business with no reputation or brand name in the console market. 3DO tried to make money off licensing the hardware to manufacturers and charging a royalty for each console sold but I believe this is partially responsible for the high hardware cost.

I think it was it was more that the manufacturers had to make a profit from the hardware.

Quote:

Maybe there was just too much hardware driving up the cost even though it was cheaper than a x86 desktop computer with similar capabilities (x86 desktop hardware was more versatile, open and expandable while 3DO was a dedicated games machine).

Yes. Just like the Amiga, you could justify spending more on a PC because it was general purpose. By the mid 90's PCs were ubiquitous not just in business but also education and the home too. So everyone had to have one. Being able to play games was just a bonus (or at least that's what you told the 'finance manager').

Quote:

CDTV was a failure at a high price too but the better priced CD32 would have been a success had C= not failed.

I agree with that too!

Quote:

Had C= survived, the CD32 may have survived longer than higher performance and superior hardware alternatives like 3DO and the Sega Saturn.

The C64 was still selling well in 1994. The Sega Mega Drive lasted until 1997. Many CD32s would get upgraded by their users to full computer status. Open hardware and software would keep the interest going long after other systems had eclipsed it. But only if Commodore had survived.

@BigD

Quote:

BigD wrote: @kolla You have to physically transfer the USB to your/PC/Mac/USB equipped Amiga since no one has hacked networking support yet! However, the 'shared' folder is visible under the AMiNIMiga if this is what you mean!?

I meant exactly what I wrote - does this involve turning off the emulator?
Or can you just pull out and insert USB sticks while emulation is running and they get automatically recognized from the Amiga side?

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

You act like there's an internal hard drive in the Mini?! The system uses the USB as its storage! You are going to hang the system if you pull out the storage while the machine is on!

Anyway this all a moot point now that network capability is being rolled out! That WOULD allow file transfer while the system is on and emulation is running! Good news huh?

Last edited by BigD on 12-Sep-2023 at 08:50 AM.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios