Overview Features Coding Performance Forum Downloads Products OrderV4 Contact

Welcome to the Apollo Forum

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



All TopicsNewsPerformanceGamesDemosApolloVampireAROSWorkbenchATARIReleases
Performance and Benchmark Results!

Vampire Future CPU Offeringspage  1 2 

Manuel Jesus

Posts 151
28 May 2016 14:52


Hi guys,
 
  In reading the threads regarding Hyperions preference for PPC dev. Would a "future" version of a Vampire offer PPC emulation? I think this would really be an equalizer. Current hardware partners are unable to produce PPC hardware at a price that resonates with Amigans.
 
  A hybrid 68k PPC core would unify and finally deliver a VIABLE ppc solition. Are current days FpGA's able to implement a PPC core?



Claudio Guglielmotti
(Apollo Team Member)
Posts 185
28 May 2016 15:27


we could ask to BigGun to stop the development of the 68000 and begin with the PPC.

Once we have the PPC we can resume apollo...
The problem is how much time needs BigGun to build from scratch a PPC ?
What we will do in the meantime?




Manuel Jesus

Posts 151
28 May 2016 15:43


I would think PPC would be best AFTER 68k is mastered no?



Philippe Flype
(Apollo Team Member)
Posts 299
28 May 2016 15:53


I don't think so. PPC is not a target. Vampire is the home of 68K lovers.


Martin Soerensen

Posts 232
28 May 2016 16:49


claudio guglielmotti wrote:
  The problem is how much time needs BigGun to build from scratch a PPC ?

Maybe 5 years to get a decent working core if we're lucky? Could also be 10 years.. But then again, these are 25 year old machine, so who cares whether we wait another 10 years? :-)


Michal Warzecha

Posts 209
28 May 2016 17:24


Why build PPC in FPGA? Can do hybrid like Phase5 did. Vampire+ some today PPC will be better option. PPC is still alive platform, no need to put PPC in FPGA


Ulf Andersson

Posts 29
28 May 2016 17:47


For the LOVE of GOD can't you see that Claudio was sarcastic! There will hopfully NEVER be a PPC FPGA core from the Apollo team! ..


Gunnar von Boehn
(Apollo Team Member)
Posts 5520
28 May 2016 18:08


Manuel Jesus wrote:

  Hi guys,
   
    In reading the threads regarding Hyperions preference for PPC dev. Would a "future" version of a Vampire offer PPC emulation?
 

 
 
There will NEVER be a Vampire with PowerPC Core.
 
We are 68k-Experts and our goal is to develop the new generation of high end 68K cores.
 
We regard Apollo-Core as superior to competing PowerPC CPUs.

We have Zero interest in developing something with PowerPC.
Also developing a good CPU core is a significant development effort.
This mean developing a good core takes several years of time.



Manuel Jesus

Posts 151
28 May 2016 18:49


Right on...If this was facebook I would hit the "like" button!


Michal Warzecha

Posts 209
28 May 2016 20:37


Of course We don't need much PPC CPU. You explain us long time ago why PPC is piece of sh.... but...
  Some time ago in some circumstances Amiga community was split on 2 parts- old, 68k and brand new PPC. We all know where PPC go (no where)but community is still split. Including PPC into Vampire will be a big chance to join all of us together. Many developers from PPC can help us creating software for 68k I think.
  Of course creating PPC inside FPGA is not right way to do that, PPC CPUs are cheap and stil available on market. Mayby, after all vampires and standalone system appear, there will be a chance to think about making system as Phase5 did long time ago? Just Vampire with PPC co-processor on board?


Lord Aga
(Apollo Team Member)
Posts 118
28 May 2016 22:37


Gunnar von Boehn wrote:

  There will NEVER be a Vampire with PowerPC Core.
 
  We are 68k-Experts and our goal is to develop the new generation of high end 68K cores.
 
  We regard Apollo-Core as superior to competing PowerPC CPUs.
 
  We have Zero interest in developing something with PowerPC.
  Also developing a good CPU core is a significant development effort.
  This mean developing a good core takes several years of time.
 

Woohoo :) Plus million for this !
Why would we need PPC at all ? Our 68K will be superior anyways.


John William

Posts 467
29 May 2016 06:08


Lord Aga wrote:

Gunnar von Boehn wrote:

  There will NEVER be a Vampire with PowerPC Core.
   
  We are 68k-Experts and our goal is to develop the new generation of high end 68K cores.
   
  We regard Apollo-Core as superior to competing PowerPC CPUs.
 
  We have Zero interest in developing something with PowerPC.
  Also developing a good CPU core is a significant development effort.
  This mean developing a good core takes several years of time.
 
 

 
  Woohoo :) Plus million for this !
  Why would we need PPC at all ? Our 68K will be superior anyways.

I agree. **** PPC.



Martin Soerensen

Posts 232
29 May 2016 07:33


Perhaps someone could create a wrapper for PPC code so it would be possible to emulate a PPC on a 68k platform, integrating the API so AOS4 system-friendly stuff could run on a 68k Apollo. It might not be particularly fast, but if the Apollo core becomes fast enough, it might be usable and enable the use of various tools made for PPC systems.


Gunnar von Boehn
(Apollo Team Member)
Posts 5520
29 May 2016 07:57


On 68K many people did code in ASM.
Therefore being able to execute native 68K code is important.

 
On PowerPC coding in ASM is rare done, today people use code in C.
If you have the sources in C, then you can compile for 68k easily.
This means being able to execute PowerPC code is much less important.
 
To make this clear:
We will not build PPC hardware.

Our focus is building new 68K hardware.
Our goal is to create new 68K hardware which will be faster than a number of older PPC systems.


Alan Haynes

Posts 140
29 May 2016 14:23


You tell 'em Gunnar.
I feel for those who went down that path of PPC but they made their choice.
There is a far greater potential to bring some of the many Amiga 68k former users back to the Amiga than the few who went PPC.

All I can say is; Where is Thierry when you need him!

Cheers,
Alan from Downunder


Michal Warzecha

Posts 209
29 May 2016 19:10


Like I said, I don't care about PPC much, almost never use it on my blizzard PPC, mayby few time when  I was just curious.
I have just ask about connect normall PPC CPU into some Vampire card in future if it's not a big deal. But it was just my suggestion, nothing more. Don't bother any way


John William

Posts 467
29 May 2016 19:31


Michal Warzecha wrote:

  Like I said, I don't care about PPC much, almost never use it on my blizzard PPC, mayby few time when  I was just curious.
  I have just ask about connect normall PPC CPU into some Vampire card in future if it's not a big deal. But it was just my suggestion, nothing more. Don't bother any way
 

 
  Yar, kidden? Manuel Jesus is famous now! He is talked about in chats and all over the Amiga community :D
 


Thierry Atheist

Posts 642
30 May 2016 16:06


ahem... LOOK AT CODE BLOAT.... peepeeC? buh-BYE!!!!

Nothing to worry about, anyway, as Big Gun has commented already.

I don't have the numbers, but ppc compiled software can make as much as 5 to 10 times bigger file sizes. And we don't need none of that, now. I still remember the EndIf command in ADOS being 40 bytes!!!!

An SD card of 32 Gigabytes, in AMIGA 680x0 software could encompass ALL x86 software ever made, probably.

(Hi Alan!)


Wawa T

Posts 695
30 May 2016 17:56


Thierry Atheist wrote:

ahem... LOOK AT CODE BLOAT.... peepeeC? buh-BYE!!!!
 
  Nothing to worry about, anyway, as Big Gun has commented already.
 
  I don't have the numbers, but ppc compiled software can make as much as 5 to 10 times bigger file sizes. And we don't need none of that, now. I still remember the EndIf command in ADOS being 40 bytes!!!!
 
  An SD card of 32 Gigabytes, in AMIGA 680x0 software could encompass ALL x86 software ever made, probably.
 
  (Hi Alan!)

another bs made up to falsify the real options and put them out of proportion, even if it is basically right, i guess. the ppc code is bigger by factor 2-3 at most, because its consisnt of a lot of risc instruction replacement, in contary to simpler cisc ones. i think also aligning circumstnces are worse, and also access to register content, so binaries and access to the data may suffer altogether. all in all, ppc even though it catches up in amulation of 68k clock for clock, is actually a worse option. but ist needs to be considered realisticly, not as propagandist matter.


Alan Haynes

Posts 140
30 May 2016 23:51


I think now that AOS 4.1 on PPC emulator in Amiga Forever is available then that should be enough for the PPC people. we have to think of the majority to get this back up and running. Catering to niche markets does not help the Amiga to move forward. PPC had its chance and as far as I am concerned the developers had their opportunity to capture more of the Amiga community but they could not produce a cheaper board and what we have today is the result.
I wonder exactly how many PPC users there are out there in Amiga land compared with the 68k users that are still around and potentially would come back with the re-vamped Amiga 68k.

I think Gunnar and the team are on the right track bringing out accelerator cards for existing Amiga's and using the experience of that to iron out the bugs and allow the users to relish in the new found power. Then release a completely new stand alone Motherboard that will blow even the accelerator cards out of the water. I firmly believe that after using an original Amiga with an add-on Apollo card most users will jump at the chance to go 3 or 4 times (maybe more) faster with completely new hardware. By that time I also think PPC will just be a distant and painful (for some) memory. Everyone makes mistakes in life; PPC was the Amiga's mistake. We won't hold any grudges, right Thierry.

Cheers,
Alan from Oz
Where's my A500 Vampire 2?


posts 27page  1 2