Overview Features Instructions Performance Forum Downloads Products Reseller Contact

Welcome to the Apollo Forum

This forum is for people interested in the APOLLO CPU.
Please read the forum usage manual.
VISIT APOLLO IRC CHANNEL



All TopicsNewsPerformanceGamesDemosApolloVampireCoffinReleases
Information about the Apollo CPU and FPU.

Could a Vampire On Latest Core Emulate PPCpage  1 2 3 

Gunnar von Boehn
(Apollo Team Member)
Posts 3431
06 Jul 2018 18:15


Vojin Vidanovic wrote:

Emulating x86  at 66-100Mhz 486/P5 does not help that much either.

With such emulated x86 you can play a huge number of PC games.
There were many great PC games in the 90th.
Command & Conquer, Diablo, Fall out, Panzer General, you name it.
 




Mr Niding

Posts 267
06 Jul 2018 18:48


@Vojin Vidanovic

Gunnar made a good argument regarding emulation that would support 1990s PC performance;

To convert 1 game, 1 developer needs X time to complete it.
To convert 2 games, 1 developer needs X*2 etc

IF you improve on emulation, be it MacOS or PC-Task, you will get access to a massive library, and 1 or more developer can work together on finetuning the performance of said emulation, and make it userfriendly etc.

Ofcourse, native games are nice too, like the stream of OCS games that are currently released like Worthy, Reshoot R etc.

Imagine a PC Task or Shapeshifter that is modified to utilize all the Apollo core features, and improved as the core is improved/expanded.


Vojin Vidanovic

Posts 348
06 Jul 2018 18:48


Gunnar von Boehn wrote:

  With such emulated x86 you can play a huge number of PC games.
  There were many great PC games in the 90th.
  Command & Conquer, Diablo, Fall out, Panzer General, you name it.
 

 
  Then optimizing  a DOS Box port would be a priority. I am aware of the brighter side of DOS game library, but for the apps, I believe Linux is the easier way out. Beside Alchemy image converter and Blue Wave Reader there are not much of DOS apps to find useful compared to Amiga counterparts.
 
  DOS Box 0.74 AGA
  EXTERNAL LINK 
  DOS Box 0.5  SDL
  EXTERNAL LINK 
  Its interesting that DOSBox AGA seems to be more recent port then OS4/MOS ones.
 
  Sadly, there seems to be a lot of limits to this port
 
  Limitations:
 
  1) No CDROM support
  2) No network support
  3) No RTG support
  4) No joystick support
  5) Does not include 486/586 CPU emulation


Gunnar von Boehn
(Apollo Team Member)
Posts 3431
06 Jul 2018 18:55


Vojin Vidanovic wrote:

Then optimizing  a DOS Box port would be a priority.

PC-Task is a very good program.
Its faster than DOS BOX.

I think continue to develop PC-Task would make a lot sense.


Vojin Vidanovic

Posts 348
06 Jul 2018 19:31


Gunnar von Boehn wrote:

  PC-Task is a very good program.
  Its faster than DOS BOX.
 
  I think continue to develop PC-Task would make a lot sense.

While PC Task became free by will of author, I  am not aware of source code, nor the development.


Louis Dias

Posts 15
06 Jul 2018 20:02


A full SDL 2.0 port would work some wonders as well.  :-)


M Rickan

Posts 142
07 Jul 2018 02:49


Steve Ferrell wrote:

And if all you want to do is run OS4, there are already better options for doing so, both real and emulated.

That can be said for any platform. The value proposition argument isn't part of the equation otherwise we'd all be all be using RPIs.

The reality is, there are only very limited, expensive options for OS4 users, making that community a receptive target market.

The ability to run OS4 or MorphOS on the Vampire would make it the de facto solution. And as Vampires become equipped with more memory and faster processing capabilities, the performance will eventually surpass the legacy hardware.


Steve Ferrell

Posts 207
07 Jul 2018 05:00


m rickan wrote:

       
         
That can be said for any platform. The value proposition argument isn't part of the equation otherwise we'd all be all be using RPIs.  The ability to run OS4 or MorphOS on the Vampire would make it the de facto solution.
         

         
You've already defeated your argument by dealing in absolutes.  So no, that can't be said for any/all platforms.  And the funny thing is that all this time I thought that the original and current design goal of the Vampire was to run 68K AmigaOS, not be a de facto solution for OS4, MorphOS or any other OS. 
     
If you're looking for a de facto system to run OS4 or MorphOS then you should be talking to A-EON and pressuring them to release the Tabor and the MorphOS team about releasing their x86 version of MOS instead of twisting this project into some sort of Frankenstein's monster/Franken-puter.
         
And to continue following your logic, why don't you just talk some devs into emulating PPC on an RPI or better yet, have them port OS4 to the RPI...or even some other cheap SBC computer or FPGA since you state that the "value proposition" doesn't apply. The RPI and several other SBC and FPGA solutions are certainly more capable in terms of raw CPU/GPU power and better suited to emulate PPC than a Vampire.
     
Then on the other hand you state that OS4 solutions are scarce and expensive, so does the "value proposition" now apply?  Make up your mind.
         
Just because the options for running OS4 are scarce and expensive doesn't make it Gunnar's or the Vampire's problem.  And all this talk of emulating PPC on the Vampire is pointless anyway because there isn't a single dev willing to take up the task for obvious reasons.  Emulating PPC just for the sake of PPC emulation on the Vampire is a ludicrous reason which no dev will take up, and if the only other valid reason to emulate PPC instructions on a Vampire is to run OS4, again, no dev in his right mind would take up the task because there's no point in it. 
         
Thank God Gunnar has stuck to his guns, pun intended, and stayed focused on his design goals.  Otherwise the Vampire wouldn't even be released because it would be in a constant state of flux because of all the clowns who show up here with some really absurd ideas and propositions.
         


Vojin Vidanovic

Posts 348
07 Jul 2018 11:01


Mr Niding wrote:

    Imagine a PC Task or Shapeshifter that is modified to utilize all the Apollo core features, and improved as the core is improved/expanded.
 

 
  That would be great and some abandonware games could be added to Coffin. Sadly, optimized PC Task/Shapeshifter are highly unlikely to happen. Hope brute force of Vamp clock as FPGA progressess could make things usable.
 
 
Steve Ferrell wrote:

  Just because the options for running OS4 are scarce and expensive doesn't make it Gunnar's or the Vampire's problem.  And all this talk of emulating PPC on the Vampire is pointless anyway because there isn't a single dev willing to take up the task for obvious reasons.  Emulating PPC just for the sake of PPC emulation on the Vampire is a ludicrous reason which no dev will take up, and if the only other valid reason to emulate PPC instructions on a Vampire is to run OS4, again, no dev in his right mind would take up the task because there's no point in it.         
 

 
  Fully agreed. However, backporting some nice utils or parts of OS4 would be great, if possible at some point. If idea is that OS 3.9/Coffin/Software for Vamp advances.
 
  First step could be that Enhancer v1.3 minus RadeonHD drivers,as promised by A-EON.


Mr Niding

Posts 267
07 Jul 2018 12:53


@Vojin Vidanovic

With regards to PC-Task, Greg is trying to get hold of the author of the program, to see the chances of access to the sourcecode.

So we will see how that pans out.


Vojin Vidanovic

Posts 348
07 Jul 2018 23:04


Mr Niding wrote:

  @Vojin Vidanovic
   
    With regards to PC-Task, Greg is trying to get hold of the author of the program, to see the chances of access to the sourcecode.
   
    So we will see how that pans out.
 

 
  Seen that on Amigans.net Best of luck, hope its preserved (source) and not to complex to continue/optimize.
 
  Please note that at EAB thread, there was a guy that has his e-mail (or had in 2010) something at pacific.net.au
 
  In the same EAB thread EXTERNAL LINK 
  a) It has stated he declined to give sources
 
  b) It is stated that in 2010, an on 060-Classics DOSBox was much compatibile and faster then last PC Task 4.x


Barana V

Posts 7
17 Jul 2018 23:56


This may be of interest EXTERNAL LINK


Roy Gillotti

Posts 194
18 Jul 2018 00:41


Barana V wrote:

This may be of interest EXTERNAL LINK 

It would have been, but lately JimDrew is no longer that warm to this project for some reason.


Barana V

Posts 7
18 Jul 2018 02:04


Is its possible that its hard to aim at a moving target? Can I ask what gives you this idea? What has he said?


M Rickan

Posts 142
18 Jul 2018 02:09


Steve Ferrell wrote:

Thank God Gunnar has stuck to his guns, pun intended, and stayed focused on his design goals.  Otherwise the Vampire wouldn't even be released because it would be in a constant state of flux because of all the clowns who show up here with some really absurd ideas and propositions.         

No one has suggested that a PPC reimplementation should take priority over anything. And on a number of occasions, Gunnar and team have indicated that there is room for growth far beyond preserving legacy Amiga applications.

If you don't think that there is inherent value in appealing to the MOS and OS4 community that's perfectly fine. But do yourself and everyone else a favour and give the vitriol a rest.


Steve Ferrell

Posts 207
18 Jul 2018 02:17


m rickan wrote:

   
Steve Ferrell wrote:

    Thank God Gunnar has stuck to his guns, pun intended, and stayed focused on his design goals.  Otherwise the Vampire wouldn't even be released because it would be in a constant state of flux because of all the clowns who show up here with some really absurd ideas and propositions.         
   

   
    No one has suggested that a PPC reimplementation should take priority over anything. And on a number of occasions, Gunnar and team have indicated that there is room for growth far beyond preserving legacy Amiga applications.
   
    If you don't think that there is inherent value in appealing to the MOS and OS4 community that's perfectly fine. But do yourself and everyone else a favour and give the vitriol a rest.
   

   
Indicating that there's room for growth beyond preserving legacy Amiga applications doesn't mean that the Apollo team plans to use the Vampire to emulate PPC or that it would even make sense.  And why would they undertake such a task?  Emulating PPC doesn't bring more software to the table for anyone which is what MOS and OS4 both sorely lack.  There would have to be a compelling reason for Gunnar to even want to undertake such a task not to mention the licensing issues/hurdles he'd have to overcome since PPC processors are still being produced.  And again, there's absolutely no reason for him to do so since PPC CPUs are still being produced and can run circles around any FPGA based PPC emulation rig.  If Gunnar and the Apollo team wanted to appeal to the MOS and OS4 communities they'd be better served creating a MOS or OS4 motherboard using off-the-components and off-the-shelf PPC CPUs.  But that would require that they shake hands with the devil, AKA Hyperion, and compete with A-Eon and Acube, and they have no desire to do that.  There's also no significant demand for MOS/OS4 hardware that can justify the team's investment, either in creating an FPGA based PPC rig or a conventional PPC that can run MOS/OS4.
 
I'm disappointed that you confuse reasoning and facts with vitriol, so no, I'm not giving anything a rest.  Sorry if that disappoints you.
   


Gregthe Canuck

Posts 248
18 Jul 2018 02:54


We had a discussion on irc the other day on emulation.
 
One interesting development is the QEMU updates supporting SAM440 emulation. It is making good process: EXTERNAL LINK 
 
Long-term (I am talking years) it would be cool if future Vampire generations could run QEMU and then you get PPC, x86 or whatever floats your boat. Of course there is a lot of work involved in getting QEMU hosted on the Vampire, even more work for I presume would be some sort of Linux port, and then of course this would require a next-gen Vampire with more RAM and more muscle. Maybe the V5 could be 30% faster again than the V4?
 
Endianness wouldn't be a problem that causes performance issues hosting PPC on x86. The PPC emulation wouldn't be as difficult as there are Apollo core instructions that match PPC instructions.
 
Fun to think about, but not for the near future. And likely many complications not yet considered :)


Roy Gillotti

Posts 194
18 Jul 2018 05:27


Barana V wrote:

Is its possible that its hard to aim at a moving target? Can I ask what gives you this idea? What has he said?

He was pissed his software (FUSION/PC-Task) was being bundled in ApolloOS/Coffin without his consent, even though it was later pulled and that bundle isn't really officially supported by the Apollo core team. Then there is consequential Forum debauchery souring it further. He likely had some other issues on top of that, I would need to scour a ton of forum history on EAB and amiga.org to know it all.



Barana V

Posts 7
18 Jul 2018 09:41


IMO: Ive known the sting of being stolen off.(Im sure we all do)If he was looking for an opportunity.... He could investigate payment and simple things like software keys etc that would stop all those issues.. Maybe someone close to him could suggest it. No need to isolate(easy to fall in that trap, tho)
Wish I knew him so I could encourage him.... Anyway.


Mr Niding

Posts 267
18 Jul 2018 11:05


@Barana V

Your view that Drew should look at Vampire/Apollo as an oppurtunity was really our view as well.

But it appears he isnt intrested in it. Too bad really, but if he has a change of heart, its not hard for him to create visibility on this aspect. He can join IRC, post on this or other amiga forums or even just release a Apollo spesific version and he will get instant positive feedback.

The ball is in his court.

posts 60page  1 2 3