I agree "classic" means 68060, not PCC, classic kickstarts never were supposed to work with PPC, and it surfaced after C= was bankrupt anyways.

a "budget" 68060 accelerator for the A3/4000 would be nice, today's prices are insane for old 060 accelerators. So something similar to the A508IDE from Kipper2k, an "A4060IDE" would be a cool thing if it was a 060-based sub-200 euro card!
Things like USB and network are already available as recent and decent cards from Individual Computers, so I wouldn't focus on that.

Having a bit of reflection on the feedback so far..

- SATA isn't really necessary to begin with - CF slot is more than ample for Amiga storage.
- The PPC is something I will add on later on - it's not necessary to start off with.
- First up is the choice of a real 68k chip or FPGA.. most likely the real deal.
- Ethernet/Wifi/USB wouldn't be hard so that's going ahead
- One SO-DIMM Slot should be plenty to start with.
- Audio is something that is important to me - as a tracker from the old days I always pined for 16bit audio... so DSP is something i am going to wrestle with. At this stage I will look into 56k DSP as there are already a large existing Amiga codebase.
- RTG is a must, but as far as chipsets go - i am not quite sure what path to go down.. If i can expand on the work done with Radeon chipsets over Mediator PCI then i can't see why these drivers can't be tweaked to run a lowish spec modern AMD Radeon M series GPU.

Seems a lot less like a pipe-dream now. Time for some action :)

At first XMOS is not DSP - it can be considered as fast I/O CPU with some DSP-like instructions.
XMOS should be definitely better in handling interrupts than any modern CPU (with large pipeline).

PPC... why not x86 or ARM? Apple giveup with PPC and seem that in future it will be more and more ubiquitous architecture... Side to this there are PPC emulators on x86.

PowerVR is completely mess - no documentation, no open source drivers - even ATI/AMD is better (as such AMD APU may be nice - see this as x86 + GPU in one).

Seem that PCIe is absolute must in your idea...

From my perspective - if i need fast computer i choose PC - cheaper and more convenient...
With native x86 AROS on typical PC i see no need for special high perf new and incompatible Amiga.
For legacy/classic Amigas i prefer FPGA CPU than spending 980 Euros on MC68060 with 60 - 80MHz clock.

I guess with FPGA you are more flexible and it can run at different speeds. Some people are already working on it. With 060 chip it might be easier to assemble.

What about soldered 1-2GB RAM? Because 2GB is the max there is no need to upgrade and soldered could be more robust.

Would it be really worthy these days? Look at continuously increasing price of 060 and its availability. I think proper 060 emulation through FPGA is the future.

Anyway THIS (go to Projekte and 060 turbokarte) seems to be the latest 060 accelerator. You can start with it.

@rarefaction_ Maybe you should have a talk with the Apollo-core team.
Also a chat with Majsta woul probably be useful to you. (www.majsta.com)

I understand your desire to use a real 060 but in all honesty, you could save yourself a lot of trouble if you could build something with the Apollo-core FPGA as they are already planning to add improved graphics.
Now.. I dont really know if they are willing to share.. but it doesnt cost you anything to chat to the guys..

About the 56001/2 DSPs.. As you say.. theres already code for it.. maybe useful... but not a very good chip by today's standards. ;-)

About the AMD GPU.. isnt all driver code written for PPC Amigas and *none* for the 68k OS 3.9?
Again, the Apollo-core team will pull something interesting out of the hat one of these days (or you know... years.. ;-)
RTG is good but what would be "hack of the year" would be the SuperAGA that the elves sing about when they are drunk ;-) A fully AGA compatible but enhanced version removing a number of classical AGA bottlenecks. (Bandwidth etc)

As you can see, Im leaning towards suggesting that you should put as much as possible from your wishlist into one good FPGA.
The Vampire 600 v2 uses an Altera Cyclone III. You probably something even better.
Anyhoo... I like crazy ideas so :thums up:

Subscribing :) I hope you are successful in bringing something to the Amigaverse. I prefer an accellerated classic to a modern machine as much as possible so the more performance I can achieve with 68k, the better.

Best of luck.

Sent from Ubuntu Touch using the Forum Browser app

Hi RareFaction this is one of the greatest troll threads ever, excellent work.
Could you also design a PPC add-on for my Atari 2600 - imagine the games I could make, I'd have no trouble racing the beam then.

I'd also like a new expansion for my MegaCD 32X combo - I reckon a MIPS or ARM CPU would complement the existing 2 68000s and SH3 it already has.

Why do you trolls comment in these threads when you have nothing useful to add? Don't like the thread, don't comment, simple really!

Become a troll hunter and kill them all. :D

Trolling, disagreeing. A fine line indeed. Taking the pi**, all part of the fun really. Making it personal, now that's crossing the line.

Real 060 or FPGA? How much less money would it cost that card with FPGA?

Sorry my ignorance, but how much speed could you have with a FPGA as a expansion card for a A1200? And for games? How are they going to behave? I mean extra frame rate for a game like Frontier is always great, but would it run at same pace?

More money surely? A decent FPGA + voltage level shifters is going to cost more than a 20 yo MC68060RC50? (EDIT : Hmmm maybe not... cheapest I could find was ~€100)

Nothing at the moment as no stable 020+ core has been released yet.

The last synthetic benchmarks of the 68000 core for Apollo Pheonix were getting somewhere in the region of 40x A1200@14MHz speed.

http://www.apollo-core.com/bringup/

However Pheonix-Apollo is closed source. i.e. you'll probably have to buy one from them

I'm not in the know, not being a member of the team, there may have been a lot of development since the last announcement.

@alexh the 68RC060 is probably the worst CPU in the world if you measure performance / $$ ;-)
I estimate that my i7 4770 @4GHz is roughly a 1000 times faster yet the price is in the same ballpark. (Maybe the i7 is twice the cost)
Once you have a solid softcore of the 060 it can become better and better as you can put the softcore into a new FPGA etc... (and in the case of the Apollo project - actually improving on the original design)
Altera were recently bought by Intel. In a year or so we'll probably see die-shrinks for the bigger Cyclone FPGA's reducing cost, increasing efficiency and performance.
If you ask me, the future for 68k looks bright as long as you stick it into an FPGA. ;-)
Regarding stable 68k cores, Id say Mist, Minimig and FPGA all have pretty good CPU cores. The bugs are usually in the AGA code.

with a wishlist like in the original post, Id rather see something like a new mobo like GBA1000 but with some fpga juice acting as the aga chipset, then your not constrained by trying to stick a power 7+ module in the trapdoor (I doubt the op has even seen a power 7+ module...).

There's not even a released 020 synthesisable core let alone 060 which is many times more complex
They have the same core TG68k.C. The only open source 68k core. It started life as a 68000 core which gained a few 020 instructions. It is not a true 020.

There is no synthesisable FPU or MMU yet....

Its a long road to come before we get an 060 soft core.

Just an idle question but what's been the problem with creating a good 020 softcore. I thought 68k series was pretty well documented. So is it "just" lots of hard work, or undocumented wierd stuff, or some fundamental problem with using current/affordable FPGA?

The resulting newer DSP on 56000 core. But I better put next to 68k low cost MPC5200B (PPC 603e) and integrate with WarpOS.

https://youtu.be/oKXgiS4nXZs

http://ebay.com/itm/MPC5200CVR400B-F...-/111170594041

MPC5200B I/O System on Module

http://i62.tinypic.com/ftn04x.jpg

PPC + FPGA (68k) + 128MB DDR / 64MB Flash
+ USB 1.1, Ethernet, ATA 1.0, CAN, UART, I2S

http://i59.tinypic.com/1enh5k.jpg

:D

No need for 68k in FPGA if you have fast CPU (PPC,ARM or x86) - legacy software can be run trough emulator, new software can be native...
But anyway this is kind of dead end...

:crazy
How is this supposed to work on/in any amiga model with 68k-only code in ROM ????
It would require to boot the alien cpu from its own boot code. That is, the machine is independant, hampered by the attached amiga model. Silly!
You better stick with emulators on cheaper machines.