A500/2000 accelerator cards: still unfeasible to create new ones?

[sim085]

I think one of the most sought of stuff in the Amiga community are Accelerator Cards. Everyone wants an accelerator cards but few can justify such a high expense for such old hardware. Every now and then someone (including me) asks the question why we don’t have new accelerator cards. The answer is usually always the same; that is that it is not economically feasible to design and build such hardware because there is no market for it.

I tended to agree with this position and relied mostly on the hope that some day there was going to be some home-brew “hobby project” that would achieve this. However a few days ago I came back to amiga.org and learned about the ACA 630/25 and the ACA 1230/28. This I learnt is new hardware produced by Indivision (Jens?). Also, to me, these seem to be new designs rather then just a re-launch of some old design (although not sure on this).

Therefore at this point; is it still unfeasible to create to design and produce an accelerator card for the A500/2000 and make a profit out of it?

[tone007]

No.

[yssing]

Well it should be possible to make a simple PCB with a SMD 68k cpu, like in the A600 and click on an ACA630.

[Kronos]

Jens has allready announced (on a1k.org) that he has solid plans for doing a 020-based card for the A500.

Proposed features:
- 020 at 14MHz
- 8MB of RAM (maybe 16Bit maybe 32Bit wide)
- A600-compatible IDE-controller
- A1200-trapdoor  connector to use A1200 style accels !!

He also briefly mentioned on doing "something better" for the A2000, but no details yet.

[Digiman]

In the past there were plenty of cards which worked on ALL 68k based OCS Amigas A500,1000,1500,2000 that just slotted into the CPU socket.

I don't see why you can't take the A600 accelerator card and just re-jig the layout of the pins to the original 68000 CPU layout and just put the pins on the bottom and a 68000 socket on the top.

It would cost very little in design terms, and that just leaves manufacture. And if you design it properly so that it fits in the A1000 too (CPU socket is close to the side of the machine IIRC) then it's all viable.

The one's on ebay are all bollox '32bit FAST RAM daughter board not included' which makes them useless despite being 25mhz 68030 cards. Without FAST RAM you get no real speed increase so for me the issue is not cost it is effectiveness.

I would love to purchase 2 of these boards, one for my A1000 and one for my A2000. Many other people would instantly purchase such a thing.

I wouldn't purchase it for my A600 though, it's not Indivisions fault, it is Commodore's fault for making the A600 such a cheap piece of tat to use compared to the A1000/2000 I own

[Lockon_15]

Jens has allready announced (on a1k.org) that he has solid plans for doing a 020-based card for the A500.

Proposed features:
- 020 at 14MHz
- 8MB of RAM (maybe 16Bit maybe 32Bit wide)
- A600-compatible IDE-controller
- A1200-trapdoor  connector to use A1200 style accels !!

He also briefly mentioned on doing "something better" for the A2000, but no details yet.

Hmm, those specs are pretty low (insluting is too strong word, but you get the picture)  compared to ones found on ACA630.

I've read complete threads at EAB & A1K and there were tons of posts dealing with space/thermal management in A600. Actually, one of them had ACA630, Indivision ECS, Subway USB alltogether with modified floppy drive crammed into poor A600 case.

A500 should be left free of those constrains and for that reason it's weird that Jens plan to build something pretty unconvincing, at least from performance aspect.
Option to use A1200 trapdoor accelerators is limited only to present ACA1230 series, so it seems ACA520 is just a gateway for securing that stock of current and perhaps future ACA1230 accelerators would remain low at preferred times.
The other reason for favoring 020 might be affordable target price; new accelerator for A500 could attract quite a bit of a Amiga crowd, far more than A600 and A1200 altogether since A500 itself is still more available for purchase and a hell lot easier for maintenance. That include people owning AGA boxes with new ACA stuff which still own one of older OCS machines; then there is significant demand from present A500 audience already owning a functional A500 accelator/storage solution which could be repalced with more common and relaible HW. I find myself in that category. but only of the final product offer a decent substitution compared to present level of...satisfaction (GVP A530).

IMHO, that new A500 stuff should (like ACA630 & ACA1230 did, for sure) include most if not all best features of earlier designs (A530, Derringer, Viper A530) as:

- full blown 68030 with at least 25Mhz
- true 32-bit FastRAM set out of Z2 address space
- buffered IDE for fast solide state (CF) storage (>5MByte persec)
- IDE & RAM enabled in fallback mode, Kick1.3 compatible
- KickFlash like features are more tha welcome, but for an any price
- clockport :roflmao:



Or maybe it's too early to see where all this A500 stuff goes...man already said Indy AGA MK2 is currently top priority...

[runequester]

I imagine the intention of the 500 expansion would be primarily as a RAM expansion.

[Buzzfuzz]

I think one of the most sought of stuff in the Amiga community are Accelerator Cards. Everyone wants an accelerator cards but few can justify such a high expense for such old hardware. Every now and then someone (including me) asks the question why we don’t have new accelerator cards. The answer is usually always the same; that is that it is not economically feasible to design and build such hardware because there is no market for it.

Yeah, I would also like to see new 060 cards, cause these cost a fortune.
Though I'm lucky to own a Cyberstorm 060/PPC 50/200 with 128MB, A Blizzard 2060/50 with 128MB, a Blizzard 1260/50 with 64MB, I'd like to have more cards, but a bit lower price than I paid for these.
This shouldn't be that hard I think, but I can't build it, I'm happy to buy them though.
 

Hmm, those specs are pretty low (insluting is too strong word, but you get the picture) compared to ones found on ACA630.

Problem with the A500 is that simply not going to be used the way we use A1200's.
A500/A600/A2000/A3000 are still all OCS/ECS Amiga's.
 
I would already be happy with a 020 with 8MB and IDE, of course being auto configured and booted up.
 
Sure I would like a full blown 030 @ 50Mhz and 32 or 64MB with a full IDE port, so 44 pins and 40 pins.
But would we use it ?
 
If I want to WHDLoad, I'm better off with one of the A1200's.

[Digiman]

Who says 8mb is enough, I had 10mb and ran out of memory on Dpaint 3 anims.

Who says 14mhz is enough....try rendering Digiview stuff down from 21bit RGB sources.

No $ale here sorry.

[TheBilgeRat]

That's all good, but how about a ZorroII/III version of Voodoo3/5, and a ZorroII/III soundcard?  Then the only thing an A2000 couldn't do is run AGA specific stuff.

[Darrin]

Well a cheap 68020 with Fast RAM will allow OS3.9 to be installed and to run WHDLoad.

My first CPU card for my A1200 was an overclocked 68020 (28MHz) with an extra 4MB of RAM and that thing satisfied me for years.

[mfilos]

Who says 8mb is enough, I had 10mb and ran out of memory on Dpaint 3 anims.

Who says 14mhz is enough....try rendering Digiview stuff down from 21bit RGB sources.

No $ale here sorry.

The majority of peepz using Amiga is for memories, whdload and some progs.
For these things an 020 with 8MB is more than enough. Then again you can get an A1200 with a memory upgrade to even have AGA instead.
With ACA630 more and more peepz are happy with their A600's since it's the elite of the accelerators for stability and power.

If you go low on memory and use intensive programs go for an 040 or better 060. Since I saw you work on 1500/2000 why don't you try getting an Blizzard 2060 all this time?
Ye I know it's expensive, but not more than most of us have gave for an BPPC

[Tenacious]

Really, the old singular solutions are still available, various accel cards, ram expansions, etc.  If you want to sell me something new for an A500, it needs to be more comprehensive in its function, sort of like the old Rejuevinator card for the A1000.

How tough is it to design a card that runs a 28 Mhz 68000, 8 Megs of fast ram, an IDE interface, and one or two clock port headers.  The whole thing should connect under the existing 68000 and not interfer with the Indivision ECS.

This would make a very capable and clean A500 without multiply expansions hanging off the left hand side.  I'd certainly buy one (or more).

[Lockon_15]

Problem with the A500 is that simply not going to be used the way we use A1200's.
A500/A600/A2000/A3000 are still all OCS/ECS Amiga's.
 
I would already be happy with a 020 with 8MB and IDE, of course being auto configured and booted up.
 
Sure I would like a full blown 030 @ 50Mhz and 32 or 64MB with a full IDE port, so 44 pins and 40 pins.
But would we use it ?
 
If I want to WHDLoad, I'm better off with one of the A1200's.

I don't think it's a question how do we use our Amigas. You use what you have at your disposal. For instance, looking at the content of your signature I can only concur that A1200 is by far best model for overall usage - for you, ins pite of a fact you do own a lot of A500s (repsect . On other side, some other people might look at A500 doing a fraction of A1200 job as a challenge and thus consider A500 expansion as priority over other options.
Don't get me wrong, for a while it's pretty common to see a lot of think-tank volumes suggesting something like 'forget A500, get a A1200'; now, with a clear opportunity to make a change for A500 we're confronted with same suggestion.

Back in late '80s, I don't recollect the time Amiga users were driving HW development by such biased or perhaps limited standpoint.

[runequester]

I figured most people who use 500's use them mainly because they are the most compatible with old games.
 
If you use it for more advanced purposes, I imagine you'd better off with something else (built-in IDE or SCSI, more expansions available etc)