Opinion: PPC, 68080 and other

[hammer]

Originally Posted by Bruce Abbott

Note that 68EC080 is a made up designation created by one of the Vampire's detractors (was it you?).

The 68080 core doesn't include an 'emulation of the chipset', that's done separately in the V4 FPGA.

You say it's a 'partial' emulation of a 68k CPU, but fail to mention that the 68020, 68030, 68040 and 68060 are only 'partial' implementations too.

Doesn't have a (compatible) MMU? Neither does the 68020 (a separate chip is needed for that, and very few Amiga accelerator cards have one).

The difference is that the 68020 has the option to be linked with 68851 MMU.

Commodore's official A2620 has 68020 and 68851 MMU chips.

Commodore's official A2630 has 68030 with MMU.

Originally Posted by Bruce Abbott

Doesn't have a full FPU? Neither does any other 68k CPU. The 68020 and 68030 have none, and the 68040 and 68060 only partially implement the 68882.

The difference is that the 68020 and 68030 have the option to be linked with 68881/68882 FPU.

Commodore's official A2620 has 68020, 68851, and 68881 FPU.

Commodore's official A2630 has 68030 with MMU and 68882 FPU.

[Gorf]

Originally Posted by Thomas Richter

I had a more positive attitude if he wouldn't give me the feeling that software is worthless, and should be free and unpaid. Gunnar abuses people for the sake of his only interest, and that's not an attitude to work with in a market that small.

He certainly does not give away his own VHDL code and/or the "68080" core design for free, but is very protective about that

"rules for thee but not for me"

[grond]

Originally Posted by Gorf

He certainly does not give away his own VHDL code and/or the "68080" core design for free, but is very protective about that

"rules for thee but not for me"

Well, producing hardware needs quite a few upfront investments and you may not be able to get your investments back because somebody else was quicker to produce and sell the hardware that you designed. If you write software, there is nothing you lose by open-sourcing it other than potential profits, you "only" invested your time.

Speaking of this, nobody gives away the AmigaOS code for free either. The profits even go into the pockets of somebody else than those that do the work. This would be totally acceptable if it didn't help to keep the Amiga locked in the bad situation where it has been for decades.

But somehow I feel all this has been debated before on one or two or possibly more occasions...

[Gorf]

Originally Posted by grond

Well, producing hardware needs quite a few upfront investments and you may not be able to get your investments back because somebody else was quicker to produce and sell the hardware that you designed. If you write software, there is nothing you lose by open-sourcing it other than potential profits, you "only" invested your time.

It would of course be terrible, if the "68080" core could be used on other FPGA boards like e.g. the MISTER ... imagine all the lost time that went into the VHDL code ... ooh wait

Or god forbid someone would take it and actually produce an ASIC version

[Thomas Richter]

It's perfectly valid to sell software and hardware and to keep it closed, and willing to sell it for money. But this requires that you also respect others to do the same, and sell P96 development licenses for money, and AmigaOs licenses for money.

[Gorf]

Originally Posted by Thomas Richter

It's perfectly valid to sell software and hardware and to keep it closed, and willing to sell it for money. But this requires that you also respect others to do the same, and sell P96 development licenses for money, and AmigaOs licenses for money.

Well - depending on where the concept/idea is from and if it is really all your own work.
There are a lot of questions around the original ideas (Natami project and original Vampire - both creators are no longer part of the Apollo team) as well as the code-basis for the "68080" (Coldfire origin?)

[Samurai_Crow]

The open-source vs. closed and for-profit vs. non-profit discussion is interesting, it's also drifting off-topic. Licensing is only tangentially related.

I think PPC is dead and, if the Chinese get licensed out of high performance cores from ARM and x86, they will give RISC-V the boost it needs to start competing with the others. It's as bi-endian as PPC and ARM but the RISC-V Foundation looks like it's trying to standardize coprocessor instructions such that drivers will not be needed for graphics cores. I pushed for standardized hardware APIs since the late 90's and now it's starting to come.

[Promilus]

I think PPC is dead and, if the Chinese get licensed out of high performance cores from ARM and x86, they will give RISC-V the boost it needs to start competing with the others

And it will still be proprietary ... either ISA extension or h/w implementation. In that case you can stick to ARM...

[Samurai_Crow]

Originally Posted by Promilus

And it will still be proprietary ... either ISA extension or h/w implementation. In that case you can stick to ARM...

Just hardware implementation will be proprietary if the standard is followed to the letter. I find it odd that I just finished saying that the RISC-V Foundation was defining extensions into the ISA as optional parts of the standard and you contradict me immediately without explanation. Did you even read my post?

[hammer]

Originally Posted by Samurai_Crow

The open-source vs. closed and for-profit vs. non-profit discussion is interesting, it's also drifting off-topic. Licensing is only tangentially related.

I think PPC is dead and, if the Chinese get licensed out of high performance cores from ARM and x86, they will give RISC-V the boost it needs to start competing with the others. It's as bi-endian as PPC and ARM but the RISC-V Foundation looks like it's trying to standardize coprocessor instructions such that drivers will not be needed for graphics cores. I pushed for standardized hardware APIs since the late 90's and now it's starting to come.

The RISC-V standard specifies that you can have big-endian data (loads/stores), but unlike Aarch64, you cannot have big-endian code.

Big and bi-endianness is supported through non-standard variants; instructions are always little-endian.

X86-64v1(AMD64) ISA patents expired in 2021 i.e. 20 + 1 years. Windows 11 H2 is moving towards X86-64v2 i.e. 1st generation Core i series. PS4/XBO and PS5/XSX game consoles promoted the use of baseline AVX and AVX2 (X86-64v3) respectively.

[Promilus]

Originally Posted by Samurai_Crow

Just hardware implementation will be proprietary if the standard is followed to the letter. I find it odd that I just finished saying that the RISC-V Foundation was defining extensions into the ISA as optional parts of the standard and you contradict me immediately without explanation. Did you even read my post?

You were talking about standard RISC-V Foundation extensions...

RISC-V developers may create their own non-standard instruction set extensions. These follow the "Z" naming convention, but with "X" as the prefix

And that's what I'm talking about...

This:

Commercial RISC-V ISA core vendors verify and guarantee their own core designs and provide technical support to their paying licensees, just like vendors of proprietary ISAs, except that RISC-V customers are free to choose among a large number of commercial core suppliers implementing the same standard ISA.

In conjunction with large number of optional standard and vendor ISA extensions makes it quite a mess. Also watch relevant parts of this video if you're still thinking RISC-V is anything actually better for us, little ants of electronic world...
[ Show youtube player ]

I think not. What exactly do you expect from RISC-V for Amiga users to be beneficial versus what ARM already offers? Technologies which gives ARM their performance are proprietary. If RISC-V ever get them as well it will most likely just match up the performance level of current ARM8.3+ technologies. With price tag attached. So what exactly is so great?

[mschulz]

Originally Posted by hammer

The RISC-V standard specifies that you can have big-endian data (loads/stores), but unlike Aarch64, you cannot have big-endian code.

On AArch64 instructions are always written in little-endian even if you work in big-endian mode.

[OlafSch]

Originally Posted by Thomas Richter

It's perfectly valid to sell software and hardware and to keep it closed, and willing to sell it for money. But this requires that you also respect others to do the same, and sell P96 development licenses for money, and AmigaOs licenses for money.

Thomas, as long this party uses something like Aros that includes everything then it is perfectly legal

[clebin]

I don't think anyone has asked what the OP actually intends to the do with this computer?

It seems that a classic Amiga is off the table, so it would be useful to know what his/her expectations of a "next-gen" Amiga are.

Without wanting to put a dampener on things (well maybe a bit.. ), it might be that no hardware or OS is currently going do what they're hoping for and it'll end up in the attic within weeks. It happens.

We can all play favourite CPUs, but surely any hardware purchase is going to be driven by the software you want to run on it.

[TCD]

Originally Posted by clebin

We can all play favourite CPUs, but surely any hardware purchase is going to be driven by the software you want to run on it.

Which has been the Archilles' heel of any 'next gen' solution

[OlafSch]

Originally Posted by clebin

I don't think anyone has asked what the OP actually intends to the do with this computer?

It seems that a classic Amiga is off the table, so it would be useful to know what his/her expectations of a "next-gen" Amiga are.

Without wanting to put a dampener on things (well maybe a bit.. ), it might be that no hardware or OS is currently going do what they're hoping for and it'll end up in the attic within weeks. It happens.

We can all play favourite CPUs, but surely any hardware purchase is going to be driven by the software you want to run on it.

I think there is a booming retro market. And there is the world outside with users mostly not even knowing the brand amiga. Expectations are very different, depending on whom you ask

[clebin]

Originally Posted by OlafSch

I think there is a booming retro market. And there is the world outside with users mostly not even knowing the brand amiga. Expectations are very different, depending on whom you ask

I'm not sure if you're agreeing or disagreeing. Because I agree, it's a huge market but expectations vary, which is why I ask what they want to do with the thing.

The OP mentioned OS4/MOS/AROS but they I'm not sure if they fit into the usual retro computing scene, or at least they're at the very fringes.

If they want a vaguely modern platform to run a web browser, I'd recommend MOS. If it's running classic Amiga games, I'd recommend a classic Amiga, an A500 Mini, possibly the A600GS or a nice FPGA device like MiST or MiSTer. If it's souped-up classic Amiga, I'd recommend a PiStorm or sticking with emulation; maybe a Vampire but probably not because of price. As I say, it really depends on use-case.

[OlafSch]

Originally Posted by clebin

I'm not sure if you're agreeing or disagreeing. Because I agree, it's a huge market but expectations vary, which is why I ask what they want to do with the thing.

The OP mentioned OS4/MOS/AROS but they I'm not sure if they fit into the usual retro computing scene, or at least they're at the very fringes.

If they want a vaguely modern platform to run a web browser, I'd recommend MOS. If it's running classic Amiga games, I'd recommend a classic Amiga, an A500 Mini, possibly the A600GS or a nice FPGA device like MiST or MiSTer. If it's souped-up classic Amiga, I'd recommend a PiStorm or sticking with emulation; maybe a Vampire but probably not because of price. As I say, it really depends on use-case.

I think retro and modern are different worlds. MOS is not really modern either, as AmigaOS and even AROS that also misses some parts. Expecially drivers and modern software will always miss because of lack of developers. So current NG concepts do not really persuade me.

Retro is covered really well now. For that you do not need NG. And to play some role outside retro you would really need different concepts. So i assume we agree.

[Thomas Richter]

Originally Posted by OlafSch

So current NG concepts do not really persuade me.

Ditto here. I'm not really convinced. That said, I believe the move using Linux as basis for Aros (and thus depend on its kernel drivers) makes for some applications more sense. The end result could be something like "vamos, but with graphics and a GUI", and that is something I believe I would appreciate. Vamos has already turned out as an indisposable development aid here in my development chain. That, plus some graphics on top might come handy.


This said, such a thing would not replace my Amiga (probably nothing would), but if the end result would allow me to mix and merge some linux applications with an Amiga GUI, it would be interesting quite some other tasks (develpment for retro, not development *on* retro - noting the difference).

[Pyromania]

Originally Posted by TCD

There might be a reason for that.

Just out of curiosity: The Vampire 'strain' of Amiga add-ons is on pause it seems? Even the website didn't get updated this year.

They are way too busy selling/shipping Vampires to radically change the website which is already functional.