@amigakit

Quote:

The A600GS does not need any "soft mods"

How is what you are offering not already a "soft mod"?

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

BigD Quote:

And yet I have only run my new game Ceccenoid on my Mini thus far. I plan to buy and run Metro Siege on the Mini too! I have yet to do any coding on my real Amiga but the code runs fine under emulation. The A600GS is going sell software on its AmiStore on an ARM emulator!! Crazy!

There are some customers of hardware using Amiga emulation who invest in software and try to turn THEA500 Mini into a more general purpose Amiga like the original but I doubt many do. Downloading a few Amiga games from the internet and putting them on a USB stick is far more likely but still provides a limited benefit to the Amiga platform or expands the Amiga user base. The A600GS does make it easier to use the Amiga as a general purpose computer with AmigaOS and Ethernet support but it is still a low end 68k Amiga with all the disadvantages of JIT emulation. An OoO ARM CPU would up the performance but the cost with it for a device that is already borderline too expensive for mass appeal like THEA500 Mini would significantly reduce sales. An OoO ARM CPU removes most of the performance killing load-to-use stalls of the cheaper in-order Cortex-A53 but needs better cooling too. Most CISC designs like the in-order 68060 and mostly in-order AC68080 don't have load-to-use stalls so the core can be smaller, cheaper, lower power and cooler than an ARM OoO core which saves money and improves value. A 68060@100MHz with AGA in a 1990s Amiga looks more impressive than THEA500 Mini or A600GS from videos I've watched while a 68060+&AA+ ASIC SoC could deliver my rough estimate of 20 to 40 times the performance with a lower price, more memory and more features. The idea should be to buy an Amiga toy and get a general purpose 68k Amiga with a mind blowing increase in value over original Amigas.

pixie Quote:

So you would envision that had the500 mini had a 68k ASIC it would be any different? Because the user base was expanded already, what is missing for new software to show up? How come there's so many minis sold and no one taping that potential user base, emulated, or not. I have a mini, being emulated is the least of my concerns, if it has no new software is dead already, what difference makes a real 68k processor?

An ASIC 68k SoC is customizable so it depends on what is wanted. The memory savings compared to wasteful JIT emulation in the case of the A600GS with 4GiB of system memory to give the Amiga 1GiB of memory may be ~$15 USD reduction in the SBC price. The ARM SoC prices used for Amiga emulation are typically in the $2-$20 USD range so there is not much to be saved for in-order Cortex-A53 cores on the low performance side but an in-order 68k CPU core like the 68060 can compete with low end ARM OoO CPU cores in performance using less area/transistors and power resulting in a cost advantage and lower system cost. Even a SiFIve U74 CPU core outperforms a Cortex-A53 core with a CISC/68060 like design despite the weak RISC-V ISA. Zero cycle load-to-use latency is good for performance in general but 68k code assumes CISC zero cycle load-to-use latency.

https://en.wikichip.org/wiki/sifive/microarchitectures/7_series Quote:

0-cycle load-to-use latency (down from 1 cycle)

The high 3 cycle load-to-use latency of the Cortex-A53 likely reduces performance of unscheduled code like JIT emulation code by roughly 50% (I can post examples again if wanted). Simulations from an old paper predicted zero cycle load-to-use latency from an in-order RISC CPU has nearly the performance of an OoO RISC CPU with 2 cycle load-to-use latency.

Zero-Cycle Loads: Microarchitecture Support for Reducing Load Latency
https://ftp.cs.wisc.edu/sohi/papers/1995/micro.zcl.pdf Quote:

This result is striking when one considers the clock cycle and design time advantages typically afforded to in-order issue processors. It may be the case that for workloads where untolerated latency is dominated by data cache access latencies (as in the case of the integer benchmarks), an in-order issue design with support for zero-cycle loads may consistently out perform an out-of-order issue processor.

Zero cycle load-to-use latency CISC cores using CISC ISAs have up to twice the performance when accessing memory compared to RISC ISAs using cores with zero cycle cycle load-to-use latency like the SiFive U74 core. An in-order CISC core may be able to compete in performance with more practical OoO cores like a RPi 4 Cortex-A72 core which is likely already at least 10 times the size of an in-order Cortex-A53 core and usually times 4 cores in a SoC. The SoC price may jump from $4 USD using Cortex-A53 cores to $8 USD using Cortex-A72 cores. A $1 ASIC 68k SoC could boost 68k performance well beyond any 68k hardware so the SoC and memory savings could result in a $20-25 USD price reduction (SBC and case mass production could reduce the price more). Value is important too as adding a 3D GPU and more I/O may increase costs by a few dollars but may be worthwhile.

pixie Quote:

Where's Trevor to tap into this new wave of users instead of the attachment he has to a dead platform of a few hundreds

Right now Trevor is subsidizing an Amiga charity instead of investing in the Amiga with any chance of success.

kolla Quote:

And then, the article you use to support your claim says... Quote: We weren’t able to replicate that for the Amiga, Jeri’s off doing other marvellous things in AR and VR. We looked at doing it again but to create a bespoke solution for it would’ve been prohibitively expensive, so it is an emulation. So... it was a ridiculous idea.

RGL is a micro business likely with fewer employees than A-Eon (primarily Darren Melbourne, Paul Andrews and Chris Smith).

https://open.endole.co.uk/insight/company/09745704-retro-games-ltd Quote:

Size Micro o Turnover is under £632,000 o Under 10 employees Total Assets £1.69M

An ASIC "would’ve been prohibitively expensive" for a micro business but RGL wanted it and looked into it. The Amiga hardware is more difficult and expensive to accurately emulate and it can be upgraded more while retaining compatibility than the C64 which are likely reasons why THEC64 Mini has sold better than THEA500 Mini.

THEC64 Mini 250,000+
THEA500 Mini 80,000+

Retrogames Ltd Timeline Analysis, Sales Volume, Future list
https://youtu.be/23bYly54pUw?t=535

We have plenty of low end crap and Frankenstein 68k Amiga hardware already. THEA500 Mini pushed too far out of the below $100 USD/Euro price range for toys and didn't offer enough hardware value for non-toy customers to sell better. No AmigaOS, no Amiga branding, no Ethernet/WiFi, barely enough performance, memory and compatibility compared to existing competition, etc. THEA500 Mini value was not in the SBC but the controllers and games.

kolla Quote:

Why are you babbling so much about the chipset - it has been reimplemented a handful of times now - do you plan to have Amiga chipset on your ASIC 68k SoC which can be ridiculously cheap to produce?

Preferably, the Amiga chipset would go into the 68k SoC, if it was mature enough, but allow FPGA chipsets with an optional external FPGA as well (eFPGA blocks in the SoC would be another option). That way the 68k Amiga on one 68k core could be used for software control while the FPGA is used for other chipsets on another 68k core. In other words, use the 68k AmigaOS as a thin embedded OS instead of ARM based Linux on emulation hardware.

kolla Quote:

Actually, just three - for A500, A600 and A1200. The one for A1000 is the A500 one with a CPU socket relocation board, and the A2000 one is the A500 one with a simple adapter for the A2000 CPU expansion socket - there are quite a few of these around, and the AC ones can even be used with PiStorm. On the other hand, where is the Draken? Also - how does it compare to PiStorm? Original PiStorm (68000 socket, A500, A1000, A2000, CDTV - with or without CPU socket relocation boards) PiStorm 600, like original PiStorm, but reworked to piggy-back on the A600 SMD 68000 CPU. PiStorm32 - PiStorm for the 32bit A1200 (also works with CD32, but not "comfortably") PiStorm16 - "Real Soon Now", new PiStorm for A600, reworked with knowledge and experience gained over the years Expect a new PiStorm16 based variant for A1000/A500/A2000/CDTV as well.

The PiStorm offers more CPU performance than a FPGA core can and I expect the competition has taken a large portion of V4AC sales. The V4AC offers value in other ways including chipset enhancements and hardware features but an ASIC is the way to solve the PiStorm problem. Gunnar is unlikely to be able to afford an ASIC from V4AC sales though and he ignored potential business partners and investors when I tried so his FPGA toy will likely remain his FPGA toy.

kolla Quote:

Exactly - it is very niche. But that's really where it's at now. What are the features of your fantasy cheap 68k SoC?

Opportunities should be explored by businesses with the Amiga IP. I'm certainly not in a position to make it happen and I was unable to do anything with the AC. I have ideas but I'm flexible.

kolla Quote:

Yes, compatibility with the "high end" AGA demos, for the most part.

AGA Demos are just a way to show off the CPU performance. There are older games that didn't work well on the 68k Amiga because of lack of performance that can and are being ported to the 68k Amiga with more 68k performance. Amiga productivity software and some Amiga games benefit from more performance. It's not the 68000@7MHz OCS 512kiB Amiga hardware that is in demand but the 68060@100MHz AGA 128+ MiB, PiStorm and V4AC hardware.

Last edited by matthey on 05-Nov-2024 at 06:30 PM.

@matthey

Quote:

AGA Demos are just a way to show off the CPU performance.

There are benchmarks for that, demos are not for showing off CPU performance.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@matthey

Sorry. I meant to comment on this when you posted it. Better late than never?

Quote:

I can't think of any legitimate reason Ben would go after the "Amiga Forever" trademark for Hyperion Entertainment's own business needs. It may have been registered as a bargaining tool with Cloanto or just for revenge.

Absolutely was a bargaining tool at that time. Confirmed by multiple parties.

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@matthey

Quote:

ASIC is the way to solve the PiStorm problem

There isn’t any PiStorm problem, what are you talking about?

Also - I get the impression that most of time you mention PiStorm, you really mean Emu68. It _is_ possible to use Emu68 without a PiStorm. If the OS3 devs could manage to get rid of amiga chipset dependencies, and basic RTG could be included in kickstart, then a Raspberry Pi with Emu68 could be an OS target. And then it could potentially work even with RPi5 and up.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla
I guess the natural step would perhaps be AROS. AROS could not avoid dependencies, but over the time migrate to the ARM side of Raspberry Pi

_________________
Indigo 3D Lounge, my second home.
The Illusion of Choice | Am*ga

@thread

Back to topic.

I link to this assessment because it illustrates the confusion about shares quite clearly and the court indeed stressed that determining the shares issue was paramount.

Source

The quote from Costel Mincea in response is obviously "key" here.

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@kolla

Quote:

There isn’t any PiStorm problem, what are you talking about?

His problem is seemingly that Emu68 for RPi exists. Compounding this is that that PiStorm also exists and allows Emu68/RPi to be used to accelerate an existing Amiga to speeds far beyond any competing accelerator solution at a reasonable price.

Apparently we are all bad people and don't appreciate the 68K legacy for thinking this is anything other than some sort of heresy.

Probably part of the reason is that the MH68100 GHz ASIC doesn't exist beyond some intangible fantasy where someone has spare millions burning a hole in their pocket and doesn't give a crap about splurging it on absolute nonsense.

Wait a minute. Has he tried asking Elon Musk? How much did the silly pillock waste on hyperloop? Or twitter? Lol.

Last edited by Karlos on 15-Nov-2024 at 09:45 PM.

_________________
Doing stupid things for fun...

@Karlos

More trolling?
You know full well it's now called X.


Sorry

P.S I'll get my coat.

_________________
If your nose runs and your feet smell, you're upside down.

@TiredofLife

True... However, it was still called twitter when he splashed billions on it.

_________________
Doing stupid things for fun...

@TiredofLife

Quote:

TiredofLife wrote: You know full well it's now called X.

I'd sooner call it "The service formerly known as Twitter" than call it by that ridiculous new name.
This is the problem when certain individuals have too much money: They can indulge in things only they think are cool.

Last edited by agami on 16-Nov-2024 at 04:15 AM.

_________________
All the way, with 68k

Whenever I click on a link to that ... freak show... the browser jumps back and forth between a handful of services that clearly still have "twitter.com" in their FQDN, so... nah, it's still twitter, just with a silly "X" skin.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@Karlos

A lot of money just to hear people talk a load of nonsense.
He could have come down the pub with me and my mates and got the same result for a lot less money.



Cheers

_________________
If your nose runs and your feet smell, you're upside down.

Going off topic, but X rulez these days! One of the last big sites with free speech!

@kriz

Lol

_________________
Indigo 3D Lounge, my second home.
The Illusion of Choice | Am*ga

@matthey

Quote:

an in-order 68k CPU core like the 68060 can compete with low end ARM OoO CPU cores in performance using less area/transistors and power resulting in a cost advantage and lower system cost.

What are you basing this on?

I guess the idea of an ASIC 68K is kind of cool, but it's not very realistic. Even if it could be built, it'll be slow.

If you increase clock speed by 10X the cycle time reduces by 10X, however this also means latency goes up by 10X the clock cycles. Without all the modern mechanisms used to deal with this, the CPU will be constantly stalling. I doubt it'll get anywhere near the performance of an Arm A53, never mind Arm's OOO cores.

The 68K instruction set won't help, multi-length instructions makes instruction decoding complex and unpredictable, trying to prefetch is going to take a load of complex logic. Then you've got potential page faults in the middle of instructions, and how 68K handles exceptions is even worse.

BTW The A53 is a 12 year old processor, and it was the low end option even then. The RPi 5 is many times faster and half the price of an A500 mini.

_________________
Whyzzat?

@kriz

Quote:

kriz wrote: Going off topic, but X rulez these days! One of the last big sites with free speech!

I don't know, I kind of think AW.net is another site which supports what in your mind constitutes free speech.
I mean, you were able to post your nonsense, and I was able to call out your nonsensical post, and we're both still here.

_________________
All the way, with 68k

@pixie

Yes, that is the way. AROS x64 for PeeCees, AROS ARM on Linux ARM for everything else :D

@kolla

Some blend of AROS ARM, Emu68 and ARM optimized parts of Sys57 / A600 GS sw
might be a real AmigaOS 5 :D Using avail hardware, keeping compatibility and enabling new features.

I mean AmigaResearch OS 5.0 or 4.2 :D FINAL!! >:D

Last edited by vox on 21-Nov-2024 at 09:42 PM.
Last edited by vox on 21-Nov-2024 at 09:41 PM.

_________________
OS 3.x AROS and MOS supporter, fi di good, nothing fi di unprofessionalism. Learn it harder way!
SinclairQL and WII U lover :D
YT http://www.youtube.com/user/rasvoja

agami Quote:

I don't know, I kind of think AW.net is another site which supports what in your mind constitutes free speech. I mean, you were able to post your nonsense, and I was able to call out your nonsensical post, and we're both still here.

Disappearing posts on AW.net are censorship not free speech. Have you not noticed the disappearing posts even recently?

@matthey

disappearing posts AFAIK had to do with geen_naam asking his account to be deleted...

_________________
Indigo 3D Lounge, my second home.
The Illusion of Choice | Am*ga