Is there any point in new 68020 and 68040 boards?

General discussions or ideas about hardware.
Post Reply
kgilmore
Posts: 12
Joined: Mon Sep 01, 2025 10:36 am

Is there any point in new 68020 and 68040 boards?

Post by kgilmore »

Some time ago when searching to see if any company was building new 68020/30/40/60 processors, I found that Rochester Electronics was still making the 68020 and 68040. While the Atari market is well-served with accelerator boards and, more importantly the existing pool of electronic engineers already overcommitted with existing work, meaning that given the small size of the market there is no practical need for yet another Atari clone, would there be any point in designing:

* A 68020 board with 68881/2 socket, full 32 bit bus design for the single-tasking GEM software base. [Possibly socketed for the unobtainium 68851 though I appreciate a FPGA version of that MMU to plug in would take considerable time to make.]
* A 68040 board, again with full 32 bit bus design aimed at the FreeMiNT base

partcularly given that demos and games quite likely would not run on them? I must admit that the 68020 board intrigues me as a proof of concept of what the Mega ST could have been; a fully-32 bit single-tasking TOS computer with the capacity for greater than 4MB memory.
Atari Mega ST4, Atari Falcon030 14MB.
User avatar
alexh
Site sponsor
Site sponsor
Posts: 1501
Joined: Tue Oct 17, 2017 4:51 pm
Location: Oxfordshire

Re: Is there any point in new 68020 and 68040 boards?

Post by alexh »

For me a New Atari ST turbo card would have to have 7MHz 68000 fallback (e.g. plug the original chip into the board for fallback) there is just too much stuff on the Atari that requires exact CPU timing. Most accelerators do not have this option and taking out the accelerator and plopping back the old CPU is a PITA.

If it also supported "ALT RAM" (is that the right term?) and mass storage in 7MHz 000 mode (e.g. 10-14MB extra 16-bit RAM in 24-bit address space and IDE) that would be adventurous.

Jumperless design (some sort of switching by using the state of the keyboard/mouse during reset would be good)

There is a 3.3v 68040 so in theory you could make a Turbo card just for the remake Atari ST motherboards (e.g. H5) I don't remember if they have an FPU/MMU or not
Senior Principal ASIC Engineer - SystemVerilog, VHDL
Thalion Webshrine - http://thalion.atari.org
ST,STf,STfm,STe,MegaST,MegaSTe,Falcon060
A500+,A600,A4000/060,CD32,CDTV
Steve
Moderator
Moderator
Posts: 3416
Joined: Fri Sep 15, 2017 11:49 am

Re: Is there any point in new 68020 and 68040 boards?

Post by Steve »

I interpreted him as to be talking about an entire machine, not just an accelerator. Basically with high-level compatibility, for clean GEM apps, but without any low-level compatibility ie; no demos etc. I personally don't think that would capture any kind of market/interest.
User avatar
alexh
Site sponsor
Site sponsor
Posts: 1501
Joined: Tue Oct 17, 2017 4:51 pm
Location: Oxfordshire

Re: Is there any point in new 68020 and 68040 boards?

Post by alexh »

I really liked the shows I've been to where people brought their Raven's (actively developed Atari 68060 clone) but I found it hard find enough interest to want to make one. The price of the 060 parts were not even a factor.

There was a bit of development done on MiSTer (the FPGA platform) for Atari clones. The Atari Falcon030 core and ST cores using hybrid 680x0 that ran on the ARM chip which was interesting.
Senior Principal ASIC Engineer - SystemVerilog, VHDL
Thalion Webshrine - http://thalion.atari.org
ST,STf,STfm,STe,MegaST,MegaSTe,Falcon060
A500+,A600,A4000/060,CD32,CDTV
User avatar
agranlund
Site sponsor
Site sponsor
Posts: 1800
Joined: Sun Aug 18, 2019 10:43 pm
Location: Sweden
Contact:

Re: Is there any point in new 68020 and 68040 boards?

Post by agranlund »

kgilmore wrote: Mon Nov 16, 2026 11:25 am would there be any point in designing:
From my point of view, yes, if it brings you joy.

In the end all of this is rather 'pointless' but hobbies don't need to make sense :)

I can only speak for myself of course but the point of Raven is and was purely that -- a supply of fun for myself.
Part of the joy was the board itself, and a much bigger part the software puzzles to solve around something like that, as well as getting an almost 'infinite' supply of low level driver type code one can pick and choose to work on thanks to the ISA bus.
Practically and from a users perspective, a Falcon+CT60 probably makes more sense. Or a Vampire V4SA if you don't have a preference about FPGA vs old chips.

Personally I wouldn't have been able to approach it from the point of view of making something that pleases everyone, I think that would quickly have started becoming a job or died the feature-creep death before any board was made :)
Post Reply

Return to “HARDWARE DISCUSSIONS”