WI: No Sega 32X?

So a unification of Mars and Saturn? In the case of the Saturn part of the equation are we referring to its earlier Jupiter form or something closer to the later cost-reduced (yet NetLink Internet Modem capable) Pluto forms?

IIRC Jupiter was what project mars was before being moved, a system 32 based console was delayed and then cancelled as Megadrive was popular overseas and they wanted to take notice what other branches wanted. IIRC there was the project away that was the one evolved into Saturn.

Pluto was just a Saturn in with integrated netlink but show SEGA was hard in pushing the internet too
 
Case on point NAOMI, that was done as the arcade version of the home console plus SEGA already took a risk of a expensive CPU with mega drive that when failed in Japan, worked very well in the west.
In the case of NAOMI, the impetus for its creation was the need for something considerably cheaper than their Lockheed-Martin-powered Model 2/3, and relatively urgently. By luck, the Katana architecture was suitable for expansion. This was a markedly different situation from 1990/1991.

3DO was already demoed and have 3D capabilities, that is why 32X was done to begin with.
Very true, however the 3DO was an expensive flop (the cheapest version cost $399, as much as the Saturn did almost two years later).

You need a device that is cheap to build, not just to sell (since you can always sell cheap by loss-leading, but it's a bad practice).

Would've a ATL Saturn with a single SH-1 CPU gotten a good Daytona USA port?
Sadly, SH-1 was too underpowered. If you want a console to be released in 1993, you need to look elsewhere (ex. Motorola 68020/68030, NEC V60/V70 and such).
 
Last edited:
How would a Sega Mars with a 16 MHz NEC V60 CPU be capable of 3D games?
Not very well. The twin SH-2 in OTL 32X were 23 MHz, and the ones in the Saturn were 28.6 MHz.

Given that the ATL Mars would (should) have a single CPU (so no parallel computing headaches), you'd want a frequency at least as high as the OTL Saturn, even if the device itself is launched cca mid-late 1993.
 
Last edited:
What CPU is a good candidate for the ATL Mars?
Hmm, I think @Nivek would be more knowledgeable in this respect. I have no idea regarding computational characteristics, so I can only look at a superficial level.
Both Motorola 68030 and NEC V70 are relatively old processors, launched in 1987. So, they would be more reasonably priced come 1993. Both are full 32-bit designs. The V70 was used in some later variants of the Sega System 32/Multi 32 arcade board (in place of the earlier V60), while the 68030 had a certain amount of instruction commonality with the 68000 used in the Genesis/Mega Drive (and, again, several series of arcade boards). So both of them would be relatively familiar to Sega's engineers.
Frequency-wise, I think the 68030 edges out the V70, since the latter was only available as a 20 MHz part, whereas the 68030 was available in multiple variants, including 25 and 33 MHz. However, I'm not sure how much this would translate in a notably increase of real-world computing power.
And I have no information regarding the price difference/disparity between the two CPUs.
 
Last edited:
Frequency-wise, I think the 68030 edges out the V70, since the latter was only available as a 20 MHz part, whereas the 68030 was available in multiple variants, including 25 and 33 MHz. However, I'm not sure how much this would translate in a notably increase of real-world computing power.
Friendly reminder the motorola were out of order processor, meaning their high clockspeed was a fraud, as they took more instructions vs what the 6502 and NEC based one could do, I still yu suzuki could be the edge to keep using NEC based CPU, as they're already working on it, if anything as Suzuki was SEGA own Miyamoto, and SEGA was obsessed with 3D too
 
Friendly reminder the motorola were out of order processor, meaning their high clockspeed was a fraud, as they took more instructions vs what the 6502 and NEC based one could do, I still yu suzuki could be the edge to keep using NEC based CPU, as they're already working on it, if anything as Suzuki was SEGA own Miyamoto, and SEGA was obsessed with 3D too
Out-of-order execution isn't necessarily a bad thing. All processor designs in use today are out-of-order. I'm not a computer scientist, but apparently out-of-order designs have an advantage when various bits of the system run at different frequencies. Something about not having to wait for a specific instruction, which minimizes stalls.
 
Out-of-order execution isn't necessarily a bad thing. All processor designs in use today are out-of-order. I'm not a computer scientist, but apparently out-of-order designs have an advantage when various bits of the system run at different frequencies. Something about not having to wait for a specific instruction, which minimizes stalls.
Yeah but that is on modern PC using Apu or CPU with more developed computing languages, plus the bless of accelerators software as DirectX or Vulkan. At the time people was still writing in assembly and CPU becoming fast enough to use C language. What was doomed the Saturn was not inventing an easy coding language till 1996, too late for the west, plus out of order in assembly you waste clock cycles
 
I know this is the 3DO but i think any console earlier than Saturn would be like the 3DO in power from SEGA https://www.neogaf.com/threads/tomb-raider-has-been-ported-to-3do.1622235/ if able to run Tomb Raider, them Sega might have good times ahead
Unfortunately, the 3DO was insanely overpriced (due to all the special snowflake custom chips), plus it had a few design flaws.

My latest idea of an early CD-based 32-bit Sega console is NEC V70 CPU, GPU consisting of an embiggened standalone version of what would become the 32X VDP (the OTL 32X VDP still relied on the Genesis one) combined with the Sega Virtua Processor in a single chip, and some kind of audio chip offering a minimum of 12 channels (preferably more, like 14-16).

How early could something like that been released, at a price that wouldn't make consumers balk like they did with the 3DO, but wouldn't drain Sega with each unit sold?
 
Unfortunately, the 3DO was insanely overpriced (due to all the special snowflake custom chips)
Unironically that was the same problem doomed the Saturn, too much chips made a motherboard too expensive to fabricate that made price drops very hard to shallow only Sony loss leading and Nintendo being so efficient with the N64 allowed them to get profit quickly.


chip offering a minimum of 12 channels (preferably more, like 14-16).
Let Yamaha handled that, they were SEGA best partner.


latest idea of an early CD-based 32-bit Sega console is NEC V70 CPU, GPU consisting of an embiggened standalone version of what would become the 32X VDP
In few words a simplify Saturn with NEC, CPU,if not expensive enough I would change the SVP to a full fledge DSP with a vector sum/multiplier,( much like PS1 would do, N64 one was integrated into the microcodes) . With some good contracts, we can see a CD based Mars ( Mega drive was earth and Mega CD was luna/moon respectly)

With that way, we could see a 350$ bucks console might be capable of 3D
 
Which would be the better ideal for a ATL Sega under better management?

Option 1 - A earlier Neptune-esque standalone 32X in 1993, followed by a properly-developed 32-bit Saturn in 1994, with the ATL Saturn being followed by a Saturn 64X in 1997 - either as a add-on for the Saturn or as a standalone console, with the Saturn 64X in turn being followed by a 6th-generation DVD-based successor in 2001.
Option 2 - The Mega Drive in 1988 as OTL, followed by a earlier Saturn in 1993 that’s essentially a Neptune with a CD drive, with the ATL Saturn being followed by a earlier proto-Black Belt in 1997 and the ATL Black Belt in turn being followed by a 6th-generation DVD-based console in 2001.
Option 3 - The Mega-CD in 1991 as OTL, followed by a properly-developed Saturn in 1994, with the ATL Saturn being followed by a 6th-generation DVD-based successor in 2001.
 
Option 3 - The Mega-CD in 1991 as OTL, followed by a properly-developed Saturn in 1994, with the ATL Saturn being followed by a 6th-generation DVD-based successor in 2001.
This, once the Mega-CD was out, that would have been the dress rehearsal for their next console, if MEGA-CD fails, well there still the genesis, but a Saturn is not dragged down bizarre choices have a long life till 2000 ITTL(By 1999 the system would already be 5 years old, meaning the successor would be DVD based in 2000)
 
If Nintendo’s deal with Sony to develop the SNES-CD hadn’t fallen through, would Sega still release the 32X or at least some form of it?
 
Top