@NutsAboutAmiga

I never saw anyone distribute the Rabbit Hole laptops outside the USA. Wasted opportunity if you ask me.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@K-L

Well, for Haiku the run of users didn't happen. but still Haiku is probabyl better on with standard hardware as it would be with (expensive and powerless) custom hardware.
And look for example to RiscOS. It was virtually dead. but now with the Raspberry Pi the number pf users increased.
Still it's nt dominating the world and will not do so.
Bbut Amiga in all of its flavours won't, too.

And if you asked me what to chose: a hardware that I can easily buy for a little pocket money (some x64 computer) or a custom computer that is not only überexpensive but also wea from its processing powr and other hardware standard I know what to chose.

For "NG" (OS4, MorphOS, AROS) it's better we leave this costom hardware nonsens ASAP!
Retro is another cup of tea though.

_________________
My programs: via.bckrs.de
MorphOS user since V0.4 (2001)

@BigD

Sure. That's not the point though. The point is that the only thing all Amiga platforms have in common is the ability to run 68K binary so why not standardise it and leverage it as a CLR?

I couldn't care a less what anyone thinks the "right" Amiga platform is as an end user, but I do think the ability to write software that works on all of them without having to maintain multiple versions for every variant that comes along is the only way to proceed without having to own or loan hardware for each one. And even then maintaing it across differences is likely more effort than is reasonable for most people.

Unless all you want to do with your Amiga is twiddle Prefs all day, in which case new software isn't for you anyway.

_________________
Doing stupid things for fun...

Possible computer related meanings of "PC" are the following.

A) "Personal Computer" (small single user or individually owned computer)
B) "IBM PC" (1981 IBM model 5150 but later applied to all x86 IBM PC compatible clones)
C) "Program Counter" (current instruction pointer register of a processor)

I strongly believe Trevor was referring to option B which means most x86(-64) computers. His statement makes sense as there may still be some bias against Intel and x86(-64) by Amiga fans from it being the inferior architecture/platform that replaced the 68k Amiga. More importantly, MorphOS will almost certainly beat AmigaOS to x86(-64) hardware. AROS x86-64 is already there too. This is smart "the fox and the grapes" marketing that we didn't want the x86-64 "sour grapes" anyway. We would much rather have the "sweet grapes" of a PPC A1222 or ARM non-AmigaOne labeled hardware that may run AmigaOS 4 with Enhancer Software including graphics and 3D drivers and an AmigaOS 4 kernel replacement and semi-replacement AmigaKit developed OS since the PPC AmigaOS 4 business model is broken and unprofitable. The grapes we pay more for and get less of have to be sweeter though!

@Karlos

The Vampire and Gunnar's way trumped everything else as far as I can see because he just got on and made decisions and implemented them. AMMX is now a thing will real world games so the fork has already happened. X86 pandering is not going to force more standardised 68k development.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@BigD

Quote:

I never saw anyone distribute the Rabbit Hole laptops outside the USA. Wasted opportunity if you ask me.

I see it as the product anyone can make and hardly anyone wants.

BigD Quote:

The Vampire and Gunnar's way trumped everything else as far as I can see because he just got on and made decisions and implemented them. AMMX is now a thing will real world games so the fork has already happened. X86 pandering is not going to force more standardised 68k development.

Bring in professional CPU architects and create a 1-2GHz 68060 like successor ASIC and the Vampire will be a thing of the past. A 68060+AGA SoC ASIC Amiga could be mass produced for roughly $1, at least before pandemic disruptions. Of course we want enhancements which would add to the cost but it could still be only a few dollars. Even a 1-2GHz 68070, AA++ chipset and an Innovative Technologies hybrid raytracing GPU could be produced for a few more dollars. Development and licensing costs start to creep up though. I think more Amiga users would rather have a 68060@100MHz than an Apollo Core although the Apollo Core is included with SAGA which adds value and AMMX provides a performance boost for a few programs. Many programs that use AMMX also work on the 68060 without using SIMD code.

@Rob

I see there two branches who both have a future...

"traditional amiga" based on 68k. It is obvious if you look at events like amiga 37. This will grow in future.

second branch I would call modernized amiga. That is more than current NG and needs both modernized OS and new software running on up to date hardware, That currently not exists but hopefully will come and bring new additional users.

@matthey

you already have mass produced processors + gpu

what will bring your concept that is better than all existing solutions? For me 68k is part of the growing retro market. There is nothing wrong with that, amiga 37 was the biggest amiga event for many years and it was 99% 68k. The market has still potential to grow. But I do not see how it can beat competition on hardware level.

@K-L

Quote:

As if porting AmigaOS 4.x or MorphOS on any "standard" Intel/AMD platform would help having more users. It did not work with Haiku and it did not work with AROS, those two being excellent for what they can do. Why the heck would it bring more people on board?

I do kinda agree but also disagree, Aros is a great platform and it’s a shame not more people use it and help with development.

But the brand still means something to some, it why there still courtroom fights over it.

There are hundreds of cheap low cost arm computers that come out of china, just look at Ali express, some are even pre-setup to be retro machines. You can get them as cheap as £30. And there the Pi boards that are pretty easy to turn into retro box’s,

https://m.aliexpress.com/wholesale/retro+game+console.html?osf=auto_suggest&spm=a2g0n.productlist.header.0

But as we seen, get the Amiga brand, do a nice custom case and custom gui and the brand gets you a lot more interest, sales and can command a much higher price.

So I wouldn’t underestimate the brand.

_________________
AmigaNG, YouTube, LeaveReality Studio

same topic every year and same answer! AROS has been working better and better every year on PCs for years now! Heck Amikit emulation is running 68k AmigaOS on any PC too. You really think that there will be that many new OS4.1 users if it’s available on any PC?l when AROS and emulation have been around for years and years??

_________________
____________________________
c64-2sids, A1000, A1200T-060@50(finally working!),A4000-CSMKIII
! My Master Miggies- Amiga 1000 & AmigaOne X1000 !
mancave-ramblings
X1000 I BELIEVE

OlafS25 Quote:

you already have mass produced processors + gpu

Are you referring to the nearly 3 decade old mass produced 68060 design and chip process? What other newer 68k hardware is mass produced?

OlafS25 Quote:

what will bring your concept that is better than all existing solutions? For me 68k is part of the growing retro market. There is nothing wrong with that, amiga 37 was the biggest amiga event for many years and it was 99% 68k. The market has still potential to grow. But I do not see how it can beat competition on hardware level.

I believe a modernized 68060 CPU design can be one of the highest integer performance and easiest to use in-order CPUs on the market. ARM has this huge market with the 32 bit Cortex-A7 and 64 bit Cortex-A53 because these cores are significantly smaller area/transistors and lower power than OoO cores but their performance is weak. x86(-64) CPUs are too big and use too much power to compete here. The 68k is a much more efficient memory munching monster as demonstrated by the 68060. The Cortex-A7 is easy prey at 1.90DMIPS/MHz. I don't think a modernized 68060 would have any problems achieving that kind of performance and is much easier to schedule for (does great with unscheduled code). Perhaps the most popular 32/64 bit CPU core in the world, the Cortex-A53 at 2.03DMIPS/MHz is more difficult prey but likely achievable without major changes to the 68060 design. Achieving superior performance with a smaller area in-order core also provides a cost advantage and the excellent code density of the 68k saving instruction caches makes up for the CISC core size increase compared to RISC cores. Do not underestimate the performance of CISC just because it is using an old process. The 68060 is officially (from Motorola) 1.56DMIPS/MHz with poor compilers while the also in-order Pentium P54C has been measured at 1.88DMIPS/MHz which is about the same as the Cortex-A7. Unofficially, the integer performance of the 68060 may be better than the Pentium. I measured 40% better integer performance of the 68060 compared to the Pentium at the same clock speed with the ByteMark benchmark and still without instruction scheduling. The 68060 performance of 1.56DMIPS/MHz would be 2.18DMIPS/MHz if increased by 40% and 40% more than the Pentium 1.88DMIPS/MHz would be 2.75DMIPS/MHz. The Pentium P54C benchmarked had high performance memory but that is the thing. CISC doesn't have the memory bottleneck that RISC has and just needs to be uncorked. These new 68060@100+MHz boards have increased internal CPU cache performance (600MiB/s @50MHz becomes 1200MiB/s @100MHz) and memory performance is up to 90MiB/s where a CSMKIII was around 50MiB/s. There may be Amiga users with rev6 68060s in newer accelerators that already have better DMIPS/MHz performance than the Cortex-A7 and Cortex-A53 (they have 32kiB L1 I+D caches and L2 caches too). The problem is the low clock speed on an old process means the weak ARM processors are better performance despite having weak performance/MHz. The Apollo core has the same problem of low performance because of low clock speed despite good performance/MHz. We need a higher clocked 68k CPU to be more competitive but then we can go after those weak in-order ARM CPU cores. This requires an ASIC and investment.

Pentium P54C measured at 1.88DMIPS/MHz
https://archive.ph/20130205075133/http://www.tomshardware.com/charts/cpu-charts-2004/Sandra-CPU-Dhrystone,449.html

@BigD

Quote:

Well if it makes A-EON's custom hardware pointless when Trevor's invested thousands of his own money into this new hardware and software amalgam then I get where he's coming from. Also, if Hyperion's AmigaOS ends up on x86 (unlikely) then it likely to go the same way as BeOS IMHO.

Trevor's comment is most likely an emotional one rather than business related. He's a self-described "Amigaholic," and has plenty of non-A-EON Amiga hardware (incl. FPGA based machines). I think you can take it at face value: he doesn't like the sound of AmigaOS on x86.

I personally prefer the idea of ARM over x86 for a different reason. Going x86 when everyone else (Incl. Apple & Microsoft) are transitioning to ARM sounds like a great way to choose yet another soon to be on the decline CPU architecture.

Quote:

I never saw anyone distribute the Rabbit Hole laptops outside the USA. Wasted opportunity if you ask me.

AFAIK, very few were interested in buying one (myself included). I'm guessing that others felt they could build their own "boot direct to WinUAE" laptop themselves. I was more interested in making a fat Mini-ITX based PowerPC laptop than buying one of those sleek machines, and that's despite seeing it running in person.

@K-L
Quote:

As if porting AmigaOS 4.x or MorphOS on any "standard" Intel/AMD platform would help having more users.

Lowering the cost of entry would indeed bring in more users. We'd already have a sizable number of new users if the A1222/Tabor had been released years ago.

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.

@Karlos

Quote:

Unless all you want to do with your Amiga is twiddle Prefs all day, in which case new software isn't for you anyway.

I nominate this for comment of the year!

Quote:

I couldn't care a less what anyone thinks the "right" Amiga platform is as an end user, but I do think the ability to write software that works on all of them without having to maintain multiple versions for every variant that comes along is the only way to proceed without having to own or loan hardware for each one. And even then maintaing it across differences is likely more effort than is reasonable for most people.

I wish more people would have this attitude instead of "only my vision of the Amiga is the real Amiga and everything else must die." I've noticed a rise in anti-Trevor hate (like wanting to destroy A-EON & push him out of the community level hate), and such behaviour serves nobody. I understand that some people don't like AmigaOS 4, but trying to destroy it won't help your own preferred Amiga future. You're not going to get a single extra person to "your cause" by pissing on their work and bullying them.

Having some kind of interoperability standard would make some level of cooperation between the diverging OS variants possible, and everyone here would win from that.

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.

@K-L

Quote:

As if porting AmigaOS 4.x or MorphOS on any "standard" Intel/AMD platform would help having more users.

Um, it would have at least one more user: Me.

What has kept me, and I'm sure many others from getting into the AmigaOS 4 ecosystem, is the price/performance gap of the PowerPC hardware. It may not draw in users from other non Amiga-oid platforms, but many remaining Amigans have been waiting/wanting for more affordable performant hardware.

Just ask yourself how many more X5000 boards (AmigaOS 4 licenses) would have been sold if it were at the price of the Sam460LE?

Now halve that price again and you're likely to double the number again.
And I'm not even talking about only commodity x86 components. ARM would also do the trick:
https://www.firefly.store/goods.php?id=161 - Starting from $459 USD


Quote:

Now, I don't see the point loosing time porting MorphOS or AmigaOS 4.x to x86 (or ARM by the way), there won't be more users (when I say users, I mean real users, people using it on a daily basis).

But you are right. That proverbial ship has sailed many years ago.

I think our best way forward is by taking a step back and revisiting the Amithlon concept.

_________________
All the way, with 68k

agami Quote:

What has kept me, and I'm sure many others from getting into the AmigaOS 4 ecosystem, is the price/performance gap of the PowerPC hardware. It may not draw in users from other non Amiga-oid platforms, but many remaining Amigans have been waiting/wanting for more affordable performant hardware. Just ask yourself how many more X5000 boards (AmigaOS 4 licenses) would have been sold if it were at the price of the Sam460LE? Now halve that price again and you're likely to double the number again. And I'm not even talking about only commodity x86 components. ARM would also do the trick: https://www.firefly.store/goods.php?id=161 - Starting from $459 USD

That SBC has very nice features for high end embedded use but is the Cortex-A76@2.4GHz competitive enough considering the price?


https://www.anandtech.com/show/15776/mediatek-announces-dimensity-1000-soc

MorphOS would have x86-64 systems with more than twice the performance. Is the ARM is cheaper and has much lower power argument good enough for a desktop? The SBC is not cheap compared to a RPi 4 which has good performance/$ despite several generations older CPU (ARM Cortex-A72 to Cortex-A73 to Cortex-A75 to Cortex-A76). The RPi 4 has most of the features wanted for basic general purpose use except for a better GPU and SATA due to those not being included in a smart phone SoC. Those features aren't too much more expensive to add to a SoC though. You get what someone else wanted when using an off the shelf SoC. Custom SoCs are relatively cheap for high volume markets and the features are perfect. The only problem is that as soon as using ARM there is no Amiga high volume market because it is the wrong CPU for a retro Amiga where the 68k is the heart and the chipset the soul of the Amiga. It would cost more to develop a custom 68k Amiga SoC but then the uniqueness would bring a premium compared to the saturated ARM SoC market.

Last edited by matthey on 25-Oct-2022 at 03:38 AM.

@matthey.

A little bit off-topic, but about efficiency: Arm vs RISC-V? Which One Is The Most Efficient?

Before someone is asking for RISC-V...

@matthey

Quote:

matthey wrote: I believe a modernized 68060 CPU design can be one of the highest integer performance and easiest to use in-order CPUs on the market.

I believe I can fly.....


In reality bringing 68k to a level where it would be a viable alternative in any market would cost far more than anybody sensible would spend in fixing a problem that does not exist.


The whole "68k is booming" is that same delusion that we have seen at the start of "NG" and had been active even in the years before.
Sure there is a market but it is limited and once you saturated that small vocal group sales will collapse.
That is true wether you are offering a 500€ (+/-) system at the performance of a 20+ year old PC or if you offer a sub 50€ retro consoles.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Hypex: When you cut out this sentence without context it might sound strange.

But I was there live and in fact Trevor said nothing more than that he would prefer porting AOS4 to ARM. So I don't understand the "stir". When you know Trevor a little bit you know that he is a funny guy (in a positive way). So what?

_________________
AmigaOne X5000/40, 2.2 Ghz, 4 GB RAM, Radeon R9 280X, M-Audio Revolution 5.1, 240 GB SSD

@Kronos

To be blunt, there's far more interest today in 68K than there is PowerPC. And while I'm sure that those running OS4 on gold plated niche hardware or MorphOS on the best of apples older HW have no regrets, it doesn't make it an attractive prospect except to a diminishingly small number of users. I would hazard a guess that neither platform has had any significant influx of new users in the last decade, just the same people updating whenever a new release comes out.

Last edited by Karlos on 22-Oct-2022 at 09:08 AM.

_________________
Doing stupid things for fun...