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
Documentation about the Vampire hardware

V4SA Schematicspage  1 2 

Noki Runner

Posts 7
05 Oct 2020 10:36


Hello this is my first post here, I signed up on purpose just because I wanted to make my intuitions explicit.
 
  You should go as much opensource - open hardware as possible.
  and build services around these open hardware and software. These are the best times ever.
 
  However, you will always remain niche if you do not, the reality of hardware and software computing has grown exponentially in power, that it will still be difficult to make it attractive beyond a certain niche of enthusiasts and nostalgics.
 
  You have to take for example what the Arduino guys did. by releasing schematics of their micro controller, and then people started playing with it on the most varied objects of electronics and robotics, a market was created that in 10 years has become gigantic and exponential. We would not have 3D printers with solid software and firmware that make us land on the future without this.
  Robotics too, start learning with it ..
 
  Then there is youtube which today makes everything in the media easily expandable in enthusiasm in a short time.
 
  FPGAs make corrections and fast improvements possible on both software and hardware sides.
 
  AROS-Apollo OS, is already a starting point for this solid mix that takes the past back into the future
 
  Then there is the open standard ISA Risc-V to which you should connect as a bridge for a total and open compatibility and favor as much as possible the creation of "new modules" on it.
 
  What wonderful times, opening the heart to the right point, could set off a perfect storm that could become an avalanche that could break any dam. Don't miss this opportunity. Look big and far on the crest of the wave.


Gunnar von Boehn
(Apollo Team Member)
Posts 5487
05 Oct 2020 12:12


noki runner wrote:

  Then there is the open standard ISA Risc-V to
 

 
What made Amiga and Atari great was the 68000 CPU.
 
If you think using another much "worse" CPU  would makes any sense for us - then you misunderstand the goal of this project.


Noki Runner

Posts 7
05 Oct 2020 14:30


Sorry, I have not misunderstanding the goal of the project, I was not referring to build or using a different cpu, but to use RISC-V ISA as an open standard of forward-looking istruction set as a communication compatibility layer among the peripherals that will come (something like what is PCI express for the world of x86 architecture, sorry I don't know how to better describe it.), because surely will be full of these hardware modules (because free from restrictions and royalties) of any type based on RISC-V ISA in the near future.
EXTERNAL LINK


Captain Zalo

Posts 71
05 Oct 2020 18:19


Noki, I think you missed the point totally.
  The goal of the project is to make a 68k compliant processor and amiga chipset in FPGA, not a general SBC for tinkerers.
 
  Of course it would be fun with a Risc-v core and a full compliant ISA, but it won't run AmigaOS or Atari TOS without heavy modification in software. Designing a Risc-v solution for Vampire cards would mean a lot of effort down the drain instead of spending it on refining the core for the primary audience - The die-hard Amiga fans. If multiple general purpose fpga cores is your cup of tea, go for a MISTer or equivalent, not a Vampire.


Noki Runner

Posts 7
05 Oct 2020 20:36


Hi Captain Zalo, I think I understand very well what the mission of the vampire project is, I would like to clarify, this list I made is not something that must be included at any cost, I have just made a list of opportunities available that the world today offers, and this of the RISC-V ISA (which is not a cpu or similar) but an open ISA standard that will help the most varied peripherals to communicate with each other, which is trying to gain a foothold, and given the numbers and adhesions that are doubling from year to year, I think it is simply an opportunity and "a help in growth" that should not be underestimated for the next few years ....
  Who knows that one day a Vampire himself could become some sort of compatible "RISC-V ISA" module for something else, I think of edge computers, and always thinking about the future maybe these "other instructions" could connect things like machine learning dedicated chips, or artificial intelligence modules or fpu or whatever that simply being new there is no need for us to have an original back compatibility with the Amiga. But by implementing them, they re-modernize into what should have been "in a parallel world back to the future in the right timeline"
 
  These views are obviously "valid" only if you are thinking about the opportunity to re-have a community that is not strongly niche but a little more flexible (as was also the original community of the 80-90s where there were wizards , but there were also those who simply loved playing on the Amigas) and why not, even trespassing beyond the concept of computers (I am thinking of robots, or microcontrollers) even just to expand the inputs and be able to invest in turn.
 
  The second important reaction that I think to see to be pushed on open source - open hardware and create services around it or even look for sponsors with companies to do things win to win or ask help of patreons etc .. is that I think is: looking for example at a Raspberry Pi that costs only 50 buck, a thriving community has quickly formed on it, which fascinates even former  ex Amiga fun users, who today are giving great strength to grow the entire community formed around it, as can be so compete with that? .... I mean, beyond a certain threshold of truly enthusiasts who are willing to spend a lot, which is obviously a niche, how else can one become really competitive on mass numbers?
  That's all.


Captain Zalo

Posts 71
05 Oct 2020 21:38


(Sorry in advance for the wall of text, but I felt obliged to flesh out a real answer.)
 
  I totally see your points and I have no issue seeing the benefits of expanding on the usefulnes of the Vampire boards. I find your logic viable and your arguments valid.
 
  Variants of your points have been discussed many times before on this board, and most have been deemed out of scope for the Apollo team. They are few and have very restricted time, meaning they want to minimize scope and functionalty to excel on the 68080 core with Amiga chipset in VHDL.
  It's not meant to be a multipurpose solution or appeal to other platforms.
 
 
 
noki runner wrote:
 
    These views are obviously "valid" only if you are thinking about the opportunity to re-have a community that is not strongly niche but a little more flexible (as was also the original community of the 80-90s where there were wizards , but there were also those who simply loved playing on the Amigas) and why not, even trespassing beyond the concept of computers (I am thinking of robots, or microcontrollers) even just to expand the inputs and be able to invest in turn.
 

 
  Flexibility was probably never on the table.
  That means they would compete with the MISTer and equivalent  products. Which they don't by keeping the whole stack from hardware design to core programming and now the ApolloOS fork in a single ecosystem designed specifically for that single purpose.
 
 
noki runner wrote:
 
  I mean, beyond a certain threshold of truly enthusiasts who are willing to spend a lot, which is obviously a niche, how else can one become really competitive on mass numbers?
 

 
  This is exactly the point I was trying to make (and clearly failed at) in my earlier post - Mass numbers and competitiveness seems to not even be considered, and I doubt it has ever been or will be in the near-ish future.
  The Apollo team are enthusiasts feeding a very small niche of enthusiasts with an extremely niche product for a single purpose - Keeping the Amiga alive for a very small, but dedicated community. To apply a fitting meme; "- This is the way."
 
  I highly respect the dedication and genuine purebred ideas that has driven the Apollo team over the years. I've been following the board for many years, and even though I find most of the community stagnant and at times quite toxic, I find the purity, honesty and clear purpose of the Apollo team and their project highly appealing.
  You may call me biased, as I'm going to throw my money at a V1200 when it eventually comes available to me. Not because I need it, but because I find the idea of putting a nitro-powered pacemaker in my frankenstein A1200 is an amusing thought. And I want to put some money back into the effort of keeping the pure and original Amiga platform alive-ish.
 
  Will it serve me better than any other computer I have in the house? Highly unlikely. Even my cellphone on power save mode outpaces the vampire and runs circles around it when it comes to usefulness. Hell - My Pi4 with Amiberry takes a dump on any real Amiga when it comes to productivity and modern day appliances. But that's not the point, is it?
 
  Lot's of love.


Ron Valen
(Needs Verification)
Posts 12/ 2
05 Oct 2020 23:47


noki runner wrote:

  Sorry, I have not misunderstanding the goal of the project, I was not referring to build or using a different cpu, but to use RISC-V ISA as an open standard of forward-looking istruction set as a communication compatibility layer among the peripherals that will come (something like what is PCI express for the world of x86 architecture, sorry I don't know how to better describe it.), because surely will be full of these hardware modules (because free from restrictions and royalties) of any type based on RISC-V ISA in the near future.
 

  The instruction set is only part of the solution when microarchitecture's implementation has a major performance impact.
 
  EXTERNAL LINK  Bad RISC-V FPGA showing. LOL
 
  EXTERNAL LINK  Western Digitalís SweRV (RISC-V)core's Coremark/Mhz yields about obsolete ARM Cortex A15. Intel Ivybridge is also obsolete.
 
  Don't assume Coremark/Mhz can linearly scale with clock speed increase. Coremark is a simple benchmark i.e. not a Quake benchmark.

 


Amiga Noob

Posts 33
06 Oct 2020 04:21


Gunnar von Boehn wrote:

 
noki runner wrote:

    Then there is the open standard ISA Risc-V to
   

   
  What made Amiga and Atari great was the 68000 CPU.
   
  If you think using another much "worse" CPU  would makes any sense for us - then you misunderstand the goal of this project.
 

 
  After watching the Vampire Italia interview and Amiga Bill show, what I'm getting at is that the Apollo team's goal is.
 
  1. Provide a more powerful/modern 68k CPU(68080).
    - As I read on other sites, there seems to be many people saying that 68k assembly is easier to code compared to other instruction sets. I guess bringing back 68k would invite more people to do low level programming.
 
  2. Provide a more powerful/modern implementation of the chipset(SAGA).
    - In terms of graphics, chipset lacks power compared to modern GPU, but at least it is easier to program(?), It also has copper which I read doesn't exist in the PC world, although there are some modern retro-computers(ZX Spectrum Next) and other niche computers that are implementing copper.
  In the interview, I heard a 3d acceleration will be added. Will this be implemented in a Heterogenous System Architecture EXTERNAL LINK ? PS4 and XboxOne are HSA and the upcoming PS5 and XboxSX are probably HSA as well. If I read correctly, HSA also makes programming easier compared to CPU + Discrete GPU.
 
  3. Provide a AmigaOS compatible OS(AROS) to circumvent legal issues surrounding AmigaOS.
    - One AmigaOS advantage is probably that it is easier to interface with compared to Windows/Linux(?).
 
  From what I read from programmers, while computers are getting more powerful, it's getting harder to interface with the OS/hardware. So I guess one good thing about the Vampire is it provides an easier/simple computer to program/interface with.
 
  A good video on how computers are getting more complicated EXTERNAL LINK   
  By the way, have you looked at some experimental ISA's beside the mainstream ones? Maybe you could get some ideas there to be implemented in next models of Vampire or the current V4. I read from one programmer that I follow that there are interesting ideas in this ISA/Computer Architecture that gives solution to ongoing complexity of programming EXTERNAL LINK An example is variable length vectors, another one is a new way of memory management(Not using Pages and TLB). I'm not a good programmer or a CPU designer at all. But just maybe there are some good ideas there that will be good for Vampire.


Gerardo G.

Posts 54
06 Oct 2020 05:02


noki runner wrote:
 
        You should go as much opensource - open hardware as possible.
        and build services around these open hardware and software. These are the best times ever.
   

   
    Can you explain why? What kind of business model are you planning to use?
   
    GitHub, as many other hardware-software pages, is full of great abandoned open source projects.
   
    Open sourcing a project do not assure a profitability, quality or faster development. Usually happens just the opposite: fragmentation, incompatibility, abandoning...
   
    As an example about what could happen to Apollo / Vampire in case of open sourcing it, you can check dozens of incredible developments related to the re-implementation of old computers and gaming devices. ZX-UNO as example: It was amazingly developed and open sourced it expecting growing but finally others took advantage using a better marketing on other more profitable projects.
   
    Unfortunately this is the reality: There are so many Amiga related companies just expecting an open sourcing for cloning it and taking a buck of money with no future or timeline in mind. As a result we would get another death project. An the main dev team? Tired and abandoning it due to not being profitable.
   
    Is it open source bad? No. Is it closed source bad? No... Just two different ways of managing a project, but at this point of development and profitability, there is only one way to go: CLOSED SOURCE.


Noki Runner

Posts 7
06 Oct 2020 07:18


Hi Gerardo González-Trejo
   
    Let's try to explain each other, it's not that you have to go opensource at all costs because it's beautiful or because it's a religion.
      It is because in my opinion, looking at reality, with the monstrous Performance of the world of x86 and arm at ultra competitive prices, you will never have the possibility to reach a critical mass. even if the product is beautiful and perfect for real geeks, just because as nature wants they will always be few.
   
    Going all open, as I said above, I look at what the Arduino guys have done, who with theirs have created whole new markets, which is now a standard where even the big names are now trying to land and follow.
   
    The Amiga community is full of Geeks and this project is already very interesting to them. Going all open would attract other nostalgic geeks from other worlds as well, there is the possibility of reaching the critical mass of collaborators as happened to the linux kernel. and corporations could be added to it.
     
    Amiga and Commodore were well known this is the biggest advantage. With a solid base as I think it already is and with adequate openness, what I believe is the singularity of opensource projects could arrive.
   
    If I were to give an example of this reaching this point of no return, I would dare to take as an example the 3D modeling and animation application, blender that has always been free and opensource, around it a community of developers, of supporters with donations, they then created business models around it by selling addons, and so on, and today it is so loved that there are also large corporations to sponsor it ... look at the numbers for yourself: EXTERNAL LINK they collect 100,000 euros a month, and can afford to keep 20 developers on paychecks!
   
    If you remember, in the 80s and 90s, the Commodore was all an apparatus of enthusiasts, and the entrances were probably transversal. there were the magazines dedicated to the commodore and the amiga, there were those who developed the software, who the games, there was probably a lesser known landing place in the industry, there were games and yes ..the piracy ...and floppy disk manufacturers .... in short there was a whole community growing and expanding around it ...
   
    Many probably have been inspired by taking for example those times and the formation of those communities .. certainly I would see what Google has done with android and with the playstore and with youtube and with who knows how much else that is not so evident, if you see, adapting "those commodore examples" to modern times with the internet.
   
    It 's true I have proposed to join the "RISC-V ISA" standard .. but if I have to be honest I would not have despised at all an "Apollo Core ISA Open Standard" ...
   
    The world of laws and finance have killed Commodore and above all the community, you should create your own laws and finance and apply them to kill them.
   
    It is above all intelligence that is well applied and directed.
   
    Everything else is a dead end, you can't face the giant Goliath directly without the right strategy.


Noki Runner

Posts 7
06 Oct 2020 08:14


Ron Valen wrote:

 
    The instruction set is only part of the solution when microarchitecture's implementation has a major performance impact.
 

 
  Exactly, but what is important here is not performance, but the formation of a Standard on which to grow, for communication between excellent or even bad architectures, is not important. What is important is that a well-practicable standard is formed without too many obstacles, the goal is to form a community of developers around it. You have to see the exponential aggregation, then potentially excellent creations can arrive on them, they will emerge and excel by their nature.


Gunnar von Boehn
(Apollo Team Member)
Posts 5487
06 Oct 2020 10:35


Noki,

I understand your opinion, and we do support the open source idea since forever but open source is not always the best option.

As you will know our goal is to revive AMIGA and to bring the best 68K CPU to the people.
We have a clear plan how to do this and we follow this plan since many years. The V4 is one part to this road. Open sourcing the V4 design makes no sense for our goals.


Noki Runner

Posts 7
06 Oct 2020 18:33


Hi Gunnar von Boehn

Gunnar von Boehn wrote:

  Open sourcing the V4 design makes no sense for our goals.
 

 
  yes, off course, not now, but in a few years, who knows, maybe there could be a series of conditions that could make going open much more convenient ..


Mark Mc Fadden

Posts 36
07 Oct 2020 01:03


Ahh, you know i didnt mean anything personal by it as the "one person getting run over by a bus" is a common agile term which im sure you aware of.
Good point on the same with Jay Miner, we'll never know about the capabilities of the Ranger.
It wasnt aimed at you specifically but more on key team members and i think its good that you clarified that the team will continue regardless though, thanks for clarifying

posts 34page  1 2