AA3000 on A1K forum - the AGA A3000

[trixster]

I’m surprised no one’s reported on this so far, it’s pretty big news!

https://www.a1k.org/forum/showthread.php?t=66906

Scrat and Matze got hold of some of the few remaining AA3000 development boards, populated them and now have working AGA A3000s. Bonkers

[NorthWay]

Was that the one with or without the DSP?

[trixster]

With. But they haven't fitted one of those yet.

[alexh]

Did it have any chips which cannot be sourced? Like the Amber 2?

[Gorf]

Originally Posted by alexh

Did it have any chips which cannot be sourced? Like the Amber 2?

no Amber on this board - no flicker fixer at all (the new AA-modes were supposed to render this unnecessary, but did not fulfill the promise as we all know now)

The chip that can not be sourced is the AT&T DSP along with its own real-time operating system, that was supposed to be accessible via new libraries from AmigaOS...

[NorthWay]

Originally Posted by Gorf

The chip that can not be sourced is the AT&T DSP along with its own real-time operating system, that was supposed to be accessible via new libraries from AmigaOS...

The chip itself can't be impossible as it was a commercially sold item and if we are pac-rats with Amiga stuff then someone else surely must have something with one inside.
It might be more interesting to know if there was supposed to be a boot ROM to go with it? (All these C= source leaks, isn't there anything from the DSP project somewhere?)

[Gorf]

The chip itself can't be impossible as it was a commercially sold item

sure: it is the same DSP that ended up in Apples "AV" Macs like the "Quadra 660AV" or "Quadra 840AV" from 1993/94

but Classic Macs are not exactly cheap...

[Jope]

Very cool. I personally wouldn't go through the trouble of hunting down the DSP though. :-)

[nexus]

better check the embedded stuff etc
might be able to use the DSP32C
these were used in telephone equipment etc aka embedded markets stuff

In the early to mid-1990s, there were several other DSP-based designs that were quite possibly inspired by the ACPA and/or Mwave. These included the AT&T VCOS/RoadRunner (AT&T DSP3210), Cardinal Technologies DSP and Adaptec AudioEdge boards (Analog Devices ADSP-2115), or Sierra Semiconductor Aria chip set (Texas Instruments TMS320C25, the same DSP the ACPA used).

[TjLaZer]

They should reverse engineer the boards to make new ones!

[nexus]

http://newit.gsu.by/projects/dspc/do.../FAQS/faq2.htm


DSP32C, DSP32xx:
32-bit floating point with 40-bits accumulator and 16/24-bit
fixed point. The 32C has three 512 x 32-bit RAMs, while the
3210 has two 1k x 32-bit RAMs and a 256x32-bit boot ROM. 32C:
Serial and parallel I/O, 3210: Serial I/O, timer,
DMA-controller. 3210 available at 50 and 66MHz

i would check into lucent telephony stuff

eg stuff like this https://www.ebay.com/itm/Avaya-Lucen...wAAOSwBdhbT-J1
https://www.ebay.com/itm/Used-Good-C...ry!33702!US!-1
has a dsp32c and 3270

https://www.edn.com/Home/PrintView?c...ItemId=4352143
DSP32C-50 MHz; 225 mA typ at 5V; 1024x32-bit RAM; 4kx32-bit ROM; 24-bit external addressing; programmable wait states (0 to 3); two DMA channels; synchronous serial port, 16-bit parallel port; no timer; 133-pin PGA/164-pin BQFP (bumpered quad flatpack). $70/$57 (10,000).
DSP3207-55/66 MHz; 150/282 mA typ at 3.3V; 2kx32-bit RAM; 256x32-bit boot ROM; 32-bit external addressing; programmable wait states (0 to 3); two DMA channels; 32-bit timer; general-purpose, 8-bit I/O port; 132-pin BQFP. $37/$42 (25,000).

DSP3210-55/66 MHz; 220/285 mA typ at 5V; 2kx32-bit RAM; 256x32-bit boot ROM; 32-bit external addressing; programmable wait states (0 to 3); two DMA channels; 32-bit timer; synchronous serial port, 16-bit parallel port; general-purpose, 8-bit I/O port; 132-pin BQFP. $42 (10,000).

[Gorf]

I doubt that any "C" variant would work, or anything other than the DSP3210:
The DSP on the board shares the same processor-bus with the 030 and the 882 and using FastRAM.

But even If you find the right chip: it would just be sitting there, doing nothing at all.

[nexus]

Originally Posted by Gorf

I doubt that any "C" variant would work, or anything other than the DSP3210:
The DSP on the board shares the same processor-bus with the 030 and the 882 and using FastRAM.

But even If you find the right chip: it would just be sitting there, doing nothing at all.

they all are 68k compatible in that family also x86
but if you cant get the 1 chip why not get something better that might be available so that way if anyone ever does do anything it would be faster

[Hewitson]

The only way the DSP would be worthwhile is if a DSP solution for other Amigas could be designed. It could perhaps be interfaced with the clockport.

[Gorf]

Also Interesting:
A "Michael Craner" was working at GVP from 94-95 and developed following:

Performed new product definition, team development, project and staff management for Commodore Amiga accelerator, video time-base-corrector (TBC), SCSI accelerator, and video effects processor. Performed PCI interface design using Flex FPGA for DSP3210 multiprocessor based digital audio workstation.

https://secure.ipnexus.com/en/directory/profiles/1235

[Dynamic_Computi]

Bummer - The A1K forums are saying they are not taking new registrations - And I can't view the forums with the article on it. Is it invitation only?

[nexus]

they are just talking about that DSP etc
not sure on the signup

[Hedeon]

Originally Posted by Dynamic_Computi

Bummer - The A1K forums are saying they are not taking new registrations - And I can't view the forums with the article on it. Is it invitation only?

Yes. I think it is due to GDPR.

[NorthWay]

Originally Posted by NorthWay

The chip itself can't be impossible

I begin to wonder if that was an optimistic assessment...I have had an eye on ebay for a long time, but not a single chip has popped up that seems to match exactly.


Does anyone know if any chips were found (and tried)?

[Valken]

I wonder can we emulate it via WinUAE? I supposed an A4000 with 030 and SCSI setup would be as close as possible? Just fantastic news for everyone.