MiniMig + PowerPC = OS4 Hardware for all

[downix]

AJCopland wrote:

downix wrote:
Not realistically, no.

:lol: well its good to get that out of the way at least! How come though? I realise that the examples I'm thinking of are commercial ones like the LinkStation or some set-top boxes, I guess they have ecomony of scale on their side.

How about the boards for PPC then? Is it feasible for us to do homebrew boards with only two layers that support a PPC chip or are they going to require more as a minimum? I ask since it gets more expensive for people to make N-layer boards doesn't it and anymore than two is instantly outside the scope of etching our own.

The sheer number of pins and the short distances that they can run makes 2-layers very difficult to manage.  The best solution would be to buy a chip with a PPC core already included, but they start at over $800 in volume.  We're in between a rock and a hard place here, sadly.  I would sooner recommend sticking it out with the original 68000 and keeping the price reasonable than take such a risk at this time.

[AeroMan]

No way to do a 2 layer board, but i think that using a embedded chip like the MPC5200 could allow to use less layers. Four or six, maybe. It is something to analyze.
Why use soldered DDRs ? Put a cheap PC memory slot

[AJCopland]

AeroMan wrote:
Why use soldered DDRs ? Put a cheap PC memory slot

People have mentioned before that it makes EMI and timing of signal on the board harder to deal with. Don't know how much harder but I was thinking of the Keep It Simple Stupid motto :-D

Andy

[freqmax]

DDR is an electrical pain. An sdram slot maybe. But DDR, nope.
A single TSSOP-54 chip can offer 64 MByte. Many Amigas didn't even get close. So it better be worth the obstacle with sockets & long nights with oscilloscope and {beep}ing.

Many things are possible, if you are willing to spend the right amount of money.

[AeroMan]

freqmax wrote:
DDR is an electrical pain. An sdram slot maybe. But DDR, nope.

Many things are possible, if you are willing to spend the right amount of money.

I agree with you. soldered chips are simpler, but the pain with the memory will be related to track lenght, in trying to avoid signal reflections, not exactly the connector itself. Soldered chips will be more reliable, for sure.
There is still another problem: the PPC is a BGA, which is a pain to solder. I can always ask my boss to pass some boards in the company's oven  :-D

[freqmax]

Actually some people have started to use bread toasters instead of pizza ovens to rewflow solder. And Minimig fits..
So it might just be doable, but you have no way of verifying solders without x-ray.

[nBit7]

Actually some people have started to use bread toasters instead of pizza ovens to rewflow solder. And Minimig fits..
So it might just be doable, but you have no way of verifying solders without x-ray.

toaster.  I haven't head of that one.  I guess it would work.  But you would need a good temp control.  You would probably be better just using a regular ovenand getting a temp sensor.

Don't really need a x-ray.  If you get the solder paste correct you only need to see the outside few rows of balls.  You can get cameras that look down the small gap.  Thats how its done by the smaller outfits.

[nBit7]

No way to do a 2 layer board, but i think that using a embedded chip like the MPC5200 could allow to use less layers. Four or six, maybe. It is something to analyze.

4,6,8 layers really is not an issue.  The minimig PCBs , can't generally be produced at home.  More layers don't add that much cost these days.  In fact it is often cheaper (at least in higher volumes) because you can make the board smaller.

If you start looking at DDR, then it is near impossible to do in less that 4layers.

[HenryCase]

downix wrote:
Not bad for 5 minutes on ebay. I just type in Sun Ultra and here they are.

I searched on eBay just like you described, plenty of cheap SPARC hardware. It's looking good.

Skyraker wrote:
Speaking of Patents... i'm sure I read recently that Gateway are about to be bought out / taken over.

I'll go find out some more.

** Bloody interesting thread btw **

*Edit* Gateway now belongs to Acer. Can we have our patents back please?

Do you know which Amiga patents Acer owns? Have AInc ever tried to obtain these patents?

downix wrote:
Just looks like AROS on Linux, but I can grab a shot for you in a bit once I recompile it.

That's a very good point, anything that will run Linux can run AROS hosted. This means we are free to choose whatever hardware we want for AROS as Linux runs on pretty much everything.

This topic seems to have split a little between OS4 solutions and AROS solutions. Ideally I'd like h/w that can run both, and as AROS runs everywhere hosted (and native given enough development time) we need to focus on OS4 h/w (that was the subject in the original post).

RW222 wrote:
You'd have a kind of dual boot machine, one option would boot to classic, just the minimig and a classic 68k OS, the other option would be to use it to bootload OS4 on the PPC hardware.

I'm sorry, I don't see the benefit of the dual booting you describe. If you can run both alongside each other (like the PPC accelerator solutions) for increased processing capability, why limit yourself? Is it just because it is easier to implement?

AJCopland wrote:
Or would it make more sense for us all to increase AmigaKits sales a little and order the Efika whilst turning the MiniMig into a PCI card for it?

That does sound like an elegant solution, but surely it would require a rewrite of OS4 (thus negating the benefits)?

Also, whatever solution we come up with should be something the Amiga community can do. Since OS4 is currently locked down in legal stuff with new hardware support looking a long way off, we need to look at getting OS4 running on new hardware without requiring a complete rewrite of the software. To me it seems we have two possible solutions to do this:

1. Reverse engineer the authorization system used by AmigaOne, put this on a chip, develop new PPC hardware around it (possibly illegal).

2. Develop new PPC compatible acceleration cards that can be used with a modified MiniMig.

If anyone can see another solution to this (OS4 emulator maybe?) please let us know.

nBit7 wrote:

No way to do a 2 layer board, but i think that using a embedded chip like the MPC5200 could allow to use less layers. Four or six, maybe. It is something to analyze.

4,6,8 layers really is not an issue.  The minimig PCBs , can't generally be produced at home.  More layers don't add that much cost these days.  In fact it is often cheaper (at least in higher volumes) because you can make the board smaller.

If you start looking at DDR, then it is near impossible to do in less that 4layers.

Just out of interest, how do multiple layer PCBs get made? Is there a homebrew way of producing this type of board? Also, does anyone know how many layers the PCBs for the old Amiga PPC accelerators used (CSPPC for example, but any information on the other boards also very welcome)? Thanks.

[jkonstan]

OS4 is PowerPC based; thus, a PowerPC solution is best practical option. A mobileGT MGT5200 from freescale would need a 6 layer PCB for the BGA at which point a Spartan3e in BGA form might as well be used. DDR soldered onto the PCB is possible if you have enough experience with DDR and SI simulation tool available. The BGAs would require a reflow oven of some sort to fabricate the PCB. A simplier/easier step would be to use off the shelf the EFIKA MGT5200 board and build a PCI MiniMIG card for use with EFIKA with a Spartan3e, some DDR/SDRAM, a VGA output, PS2 connectors (mouse/keyboard), Joystick ports, serial port, IDE, etc ...

  :-)

[HenryCase]

Could this be useful for designing a PPC based Minimig:

http://www.techonline.com/learning/webinar/199203327

We could copy the Minimig architecture within a PPC chip for 68k apps and have PPC instructions for OS4 on the same chip. Or have I got this completely wrong?

[downix]

jkonstan wrote:
OS4 is PowerPC based; thus, a PowerPC solution is best practical option.

OS4 is just OS 3.1 ported to PowerPC, so I understand, I'm sure folk with more know-how can be more clear on this.  But, being that, there is no practical reason for locking it onto PowerPC.  Infact, I'd almost say it would be smarter to keep the OS on m68k and use a fast JIT emulator to bring it to other CPU's, a la Amithlon.

[HenryCase]

Again, need some feedback on whether this would be suitable...

http://www.xilinx.com/products/silicon_solutions/fpgas/virtex/virtex4/capabilities/powerpc.htm

If it is though, the Virtex-4 may be a good solution, especially considering the price, speed and expandability of the architecture:

Price: Depends on chip used. For example, the 'XC4VLX25-10FFG668C' is 253.20USD from Digikey:
http://parts.digikey.com/1/parts/586328-ic-fpga-virtex4-24k-668fbga-xc4vlx2510ffg668c.html
If this seems too much (Spartan-3 is 25.45USD from Digikey, though you wouldn't need both chips for a PPC MiniMig), think about the prices being paid for Amiga PPC accelerators now.

Speed: 700MHz max. Plenty fast enough for Amiga users needs at the moment.

Expandability: In-built Auxiliary Processor Unit (APU) supports hardware accelerators. Since the platform would be open, we would have a system that could support our needs for a long time.

This Virtex-4 development board is rather pricey (not that I know what the average price of such a device is) at 995GBP, but does include some decent features, including built in Spartan-3 FPGA (may be optional), which would speed up the development of a PPC MiniMig:

http://www.enterpoint.co.uk/moelbryn/broaddown4.html

Any thoughts?

[HenryCase]

Been doing a bit of research into PowerPC architecture, initially to see if this Virtex-4 option is worthwhile pursuing. As far as I can tell, all PowerPC chips are compatible with each other, so the PowerPC 405 core(s) in the Virtex-4 should be able to handle all the calls made to it. The only tricky part would be to reverse engineer the control electronics from a PPC accelerator board so that we don't have to write new drivers (this might not work, in fact it might be easier just to write new drivers).

I came across a few pieces of useful information during my internet search. Firstly, found out about how PowerPC came about, how PowerPC and Power architecture are linked but not compatible. Check out this Power architecture family tree:
http://en.wikipedia.org/wiki/Image:POWERhistoryfamilydiagram.png

Found the PowerPC 405 user manual, which will be useful should anyone decide to use the Virtex-4 solution:
http://www-01.ibm.com/chips/techlib/techlib.nsf/techdocs/A07CE56994E69BFE0025731C005C9BFE/$file/ppc405S_um_01_pub.pdf

Finally, read the last paragraph of the first post here:
http://www.power.org/forums/viewtopic.php?t=712
Free SPARC and MIPS clones of PPC chips exist! Tried to find them, no luck so far, but I would imagine it's a cheaper solution than the Virtex-4 one. If we made our own SPARC or MIPS core we could even integrate the MiniMig (complete with 68K CPU processes) into one chip capable of running OS4, given enough time.

[downix]

HenryCase wrote:

Finally, read the last paragraph of the first post here:
http://www.power.org/forums/viewtopic.php?t=712
Free SPARC and MIPS clones of PPC chips exist! Tried to find them, no luck so far, but I would imagine it's a cheaper solution than the Virtex-4 one. If we made our own SPARC or MIPS core we could even integrate the MiniMig (complete with 68K CPU processes) into one chip capable of running OS4, given enough time.

No, that's not what he said.  He said that free clones of SPARC and MIPS CPU's exist.  You can find both on the opencores.org website.  SPARC is an open standard, Sun even released the code to their high-end UltraSPARC T1.  MIPS, however, is not, and true clones are liable for lawsuit due to patents.  Same with PowerPC.

[HenryCase]

downix wrote:
No, that's not what he said.  He said that free clones of SPARC and MIPS CPU's exist.  You can find both on the opencores.org website.  SPARC is an open standard, Sun even released the code to their high-end UltraSPARC T1.  MIPS, however, is not, and true clones are liable for lawsuit due to patents.  Same with PowerPC.

Damn. Well, thanks for clearing this up for me downix.

Slightly OT, I was looking around the opencores.org website yesterday, came across this device:
http://www.opencores.org/projects.cgi/web/usbhostslave/overview
If the 68k core is integrated into the MiniMig FPGA, freeing up enough I/O capacity, this seems like a cheap way of adding the much requested USB functionality to the MiniMig. Of course we'd have to write our own drivers for it.

Incidentally, wouldn't there be legal repercussions if the 68k chip was emulated in FPGA form? And could we be sued by the A500 custom chip patent holders (I doubt OCS is open source)?