Overview Features Coding ApolloOS Performance Forum Downloads Products Order Contact

Welcome to the Apollo Forum

This forum is for people interested in the APOLLO CPU.
Please read the forum usage manual.
Please visit our Apollo-Discord Server for support.



All TopicsNewsPerformanceGamesDemosApolloVampireAROSWorkbenchATARIReleases
The team will post updates and news about our project here

Poll of CPU Architecture Interest for AmigaOS

Gerardo G.

Posts 54
13 Feb 2019 07:12


Hi all!

I just saw this poll on Amigaworld:

EXTERNAL LINK 
I´m sure interest on 68K is much higher than that...

Cheers!


Nixus Minimax

Posts 416
13 Feb 2019 08:26


If people think that x86 is the way to go, why don't they all use AROS? Its performance problems don't matter on GHz CPUs...



Mr Niding

Posts 459
13 Feb 2019 09:01


I agree Nixus.
 
Yes, its quite obvious that the fastest game in town is X86, but thats unrealistic from the point of getting the Dev team to reprogram/convert from PPC to x86.
Ssolie did say that he gets enough of mainstream platform at work. When he gets home and want to have fun, he dabbles with PPC/AOS.
He finds it intresting to work with.
 
You can have your opinion regarding AOS devs communication or pace, but its hard to argue against his point regarding having fun with what basically boils down to his hobby (with some pay).
He said all of this during a Amigameet in Sacramento a few years ago.
 
So those wanting a different platform is living in a dreamworld.


Vojin Vidanovic
(Needs Verification)
Posts 1916/ 1
13 Feb 2019 09:54


Mr Niding wrote:

  So those wanting a different platform is living in a dreamworld.
 

 
  Well, Steven can laugh out any serious question, but yes,
  its pointless to have such a discussion when you have just two
  major milestones - 68k OS 3.9 as is and OS 4.1.x PPC.
 
  OK there is AROS which is cross platform, but that is other question and labor of much work meanwhile, in portability direction.
 
  In the end there once was a bit x64 AOS as AmigaXL/Amithlon and that did not save the day in terms of FURTHER OS development and APP DEVELOPMENT.
 
  Having faster AmigaOS was enabled by faster 68k and even PPC h/w
 


Ronnie Beck
(Apollo Team Member)
Posts 199
13 Feb 2019 10:43


In the back of my head, I am wondering if the Apollo Core is just a part of the "68k" or if it is a separate category.  I am also wondering if my post this question will land me some nasty comments here from puritans and terminology nazis.  :-D


Gerardo G.

Posts 54
13 Feb 2019 11:28


I think the poll is not about CPU architecture, it´s about marketing a business polarizing the poll and hiding important info to manage the game board...

  "Cui prodest scelus, is fecit"




Andy Hearn

Posts 374
13 Feb 2019 11:45


Agreed.
I think the post at the bottom of the first page sums up my thoughts on this quite nicely:-


cdimauro

Re: Poll of CPU architecture interest for AmigaOS
Posted on 29-Jan-2019 5:55:51

Currently I decided to don't vote, because some clarification is required IMO.

First, what we intend for "AmigaOS"?
The (original) Amiga o.s.? AmigaOS4? Some re-implementations (MorphOS, AROS)?
The aboves, but running on an hosted environment? Or an emulated one? Or a "virtualized" one (e.g. transparently running "AmigaOS" applications on another o.s.)? Etc.

Second, which kind of "future" do we expect?
Running legacy 68K software faster / with more features (e.g.: improved RTG/RTA. Something else)? Enhancing 68K software? Enhancing the o.s. (e.g.: new APIs), maybe to put it on par with other modern o.ses? Rewriting 68K software and or/o.s.? Making it easier to port software from other platform? Etc.
The same questions apply to AmigaOS4, MorphOS, AROS, with 68K eventually changed to the wanted CPU architecture.

Third, what benefits should the CPU architecture bring to the "AmigaOS"?

to my mind, as it stands at the moment, i'm mentaly ticking the box for team Apollo and the '080, as we can run everything that got us into the amiga (and hopefully Atari/68k mac) in the first place. With all the bells and whistles we've always wanted and been dreaming of, plus now have a dev path forward for more exciting potential, as demonstrated by the guys producing stuff that uses the power given.


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
14 Feb 2019 06:37


Andy Hearn wrote:

  i'm mentaly ticking the box for team Apollo and the '080, as we can run everything that got us into the amiga (and hopefully Atari/68k mac) in the first place. With all the bells and whistles we've always wanted and been dreaming of, plus now have a dev path forward for more exciting potential, as demonstrated by the guys producing stuff that uses the power given.
 

 
  You are right.
APOLLO 68080 offers several virtues

a) Its a real 68k CPU, which means it can very nicely be coded in ASM - in Amiga coders style.
b) Its a real 68K CPU, means it can run the old demos and games natively without emulation.
c) Its future proof and providing modern features and enhancement like Multimedia-instructions, 64bit processing, and new work in progress 3D-acceleration
d) Today even in affordable FPGAs APOLLO reaches a performance much higher than any 68k ASIC. It could be used in expensive FPGAs and would then reach even a much higher performance beating several PPC ASICs - in the future this performance increase can be expected also in future affordable FPGAs. And once the 68080 core is put into an AISC, it would rival other many CPUs like PPC/ARM/x86.


Greg Thomas

Posts 24
15 Feb 2019 06:31


Where ever we go in future, its important for now to stick with 68k core, which means 68080.
Once we get modern software we may consider switching to another CPU family, sticking as we are, or doing a hybrid of both at the same time like 68080 + G3
or perhaps 68100 + G5

Whatever the case, for now we want the hardware to change as little as possible (CPU family) while software catches up. 68080 is the way to go for the next couple of years. Maybe 64bit 68k after that as we jump to 64bit OS without 32bit memory and single core limits.


Nixus Minimax

Posts 416
15 Feb 2019 07:55


Greg Thomas wrote:
like 68080 + G3
  or perhaps 68100 + G5

The successor to the 68080 ("Oh-Eighty") is the 680A0 ("Oh-Ay-Tee")...



Mario Angelucci

Posts 11
15 Feb 2019 14:34


Agree but an asic 68080 at (in example) 800 Mhz to run what software ? Which Software Houses will venture to work on the only one 68k based computer on the world nowadays ? I like the project and still waiting for the stand alone and/or A1200 version but the future is another thing. Personally I'd like Amiga OS were natively ported to arm, although any other architecture with a future ahead is welcome :)


Philippe Flype
(Apollo Team Member)
Posts 299
16 Feb 2019 12:27


Well, anyone have his own expectations, all are respectables.
     
Imho, by essence, the whole Vampire project is not meant to switch to any other CPU architecture, modern or not.
     
Leitmotiv is "68K Rulez!".
     
This is not meant to argue that 68K is best processor ever (obviously) but to support this CPU for what it is.
     
There is still a large bunch of ASM coders that still like the 68K assembly arch, because it is nice to code, clear syntax, very good features, easy to understand. Vampire purpose is to let them enjoy coding fast 68K and/or rediscover it.
   
   
   
So remember, 68K Rulezzzz!  This is not negotiable because it is all about affect :p


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
16 Feb 2019 12:41


Mario Angelucci wrote:

Agree but an asic 68080 at (in example) 800 Mhz to run what software ?

All software!

It can natively run ALL old Software - and also ALL new software.
New Amiga software you will anyway need to compile for Amiga.
Whether you compile it for 68K or PPC or XXX is the same.

Thereby its obvious the 68K will give you the most.


posts 13