Click Here
home features news forums classifieds faqs links search
6018 members 
Amiga Q&A /  Free for All /  Emulation /  Gaming / (Latest Posts)
Login

Nickname

Password

Lost Password?

Don't have an account yet?
Register now!

Support Amigaworld.net
Your support is needed and is appreciated as Amigaworld.net is primarily dependent upon the support of its users.
Donate

Menu
Main sections
» Home
» Features
» News
» Forums
» Classifieds
» Links
» Downloads
Extras
» OS4 Zone
» IRC Network
» AmigaWorld Radio
» Newsfeed
» Top Members
» Amiga Dealers
Information
» About Us
» FAQs
» Advertise
» Polls
» Terms of Service
» Search

IRC Channel
Server: irc.amigaworld.net
Ports: 1024,5555, 6665-6669
SSL port: 6697
Channel: #Amigaworld
Channel Policy and Guidelines

Who's Online
77 crawler(s) on-line.
 24 guest(s) on-line.
 0 member(s) on-line.



You are an anonymous user.
Register Now!
 Fl@sh:  20 mins ago
 paolone:  26 mins ago
 Birbo:  28 mins ago
 AP:  30 mins ago
 Trixie:  44 mins ago
 gonegahgah:  54 mins ago
 AmigaOldskooler:  1 hr 8 mins ago
 Rob:  1 hr 12 mins ago
 olsen:  1 hr 25 mins ago
 michalsc:  1 hr 42 mins ago

/  Forum Index
   /  Amiga General Chat
      /  [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Register To Post

Goto page ( 1 | 2 Next Page )
Poll : When will brand new, FPGA-based PowerPC hardware hit the Amiga market?
Within the next year
In 2 - 3 years
In more than 3 years
 
PosterThread
Petah 
[POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 29-Aug-2021 12:14:49
#1 ]
Regular Member
Joined: 10-Mar-2003
Posts: 422
From: EU <3 ❤️

As FPGA-based hardware development progresses, more and more complex platforms are being supported by the technology's cycle-exact foundation. What used to be a challenge - mimicing circuitry such as the Motorola MC680x0 family of 32-bit CPUs - is now a walk in the park for modern FPGA-based motherboards. With the current pace, real PowerPC motherboards, Zorro and A1200/A600/A500 expansion cards should be commercially available in a matter of time allowing a wider user base to upgrade their systems to OS 4.1 and beyond.

In this AmigaWorld.net Special Poll, you are kindly but firmly asked to assess just how long that period of time is, i.e. just when will you be able to deck out your Amiga with dirt cheap, genuine PowerPC hardware based on FPGA technology?

_________________
That'll Put Marzipan In Your Pie Plate, Bingo
💻 Pro-Amiga, 🌍 Pro-Globalism, 🍅 Pro-Vegan & 🇪🇺 Pro-Joint EU Defense Intervention Initiative

 Status: Offline
Profile     Report this post  
pavlor 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 29-Aug-2021 12:31:21
#2 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9248
From: Unknown

@Petah

Well, you can get cheap PPC SoC for far less with better performance than huge and expensive FPGA needed for such a task.

Just compare FPGA used in Vampire to good old 460EX in SAM. So, it may take some time until PPC FPGA implementation is viable for our needs.

 Status: Offline
Profile     Report this post  
Rose 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 29-Aug-2021 12:47:01
#3 ]
Cult Member
Joined: 5-Nov-2009
Posts: 927
From: Unknown

@pavlor

Quote:
Well, you can get cheap PPC SoC for far less with better performance than huge and expensive FPGA needed for such a task.


This, you can get 1.2Ghz T1020 for ~80€ so not really worth of it.

 Status: Offline
Profile     Report this post  
NutsAboutAmiga 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 29-Aug-2021 13:15:39
#4 ]
Elite Member
Joined: 9-Jun-2004
Posts: 11871
From: Norway

@Petah

My guess is that FPGA PowerPC is going to be slower then Sam440 or Sam460,

https://misterfpga.org/viewtopic.php?t=1400

An FPGA + PowerPC might more interesting combination, provided that ExecSG team, Trever is interested in that.

More semi compatible hardware with CIAA/CIAB timers, blitter and Paula might useful, focusing system friendly, or close to system friendly code.

it might be possible do a WHDLOAD provided, that JIT compiler can be reallocated, better use for FPGA, or improve UAE in some way, accelerate, better timing, and or provide the IO ports.

Last edited by NutsAboutAmiga on 29-Aug-2021 at 07:45 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

 Status: Offline
Profile     Report this post  
OneTimer1 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 29-Aug-2021 16:25:48
#5 ]
Cult Member
Joined: 3-Aug-2015
Posts: 676
From: Unknown

Quote:

Petah wrote:
... technology's cycle-exact foundation.


Seems like you switched to techno babble like we have seen by Belxjander
Belxjander Perception IME (Bounty)
Belxjander Polymorph VM (Bounty)
https://amigaworld.net/modules/newbb/viewtopic.php?topic_id=35268&forum=49#796721

Quote:

Petah wrote:
when will you be able to deck out your Amiga with dirt cheap, genuine PowerPC hardware based on FPGA technology?


When hell freezes over and the AOS4 sources are open sourced.

Oh here is a link to an FPGA board running

RISC-V https://www.elektroda.com/rtvforum/topic3598849-120.html#19460186
68k AmigaOS with floppy https://www.elektroda.com/rtvforum/topic3598849-120.html#19383752



Last edited by OneTimer1 on 29-Aug-2021 at 06:19 PM.
Last edited by OneTimer1 on 29-Aug-2021 at 06:12 PM.

 Status: Offline
Profile     Report this post  
OneTimer1 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 29-Aug-2021 16:33:01
#6 ]
Cult Member
Joined: 3-Aug-2015
Posts: 676
From: Unknown

Quote:

NutsAboutAmiga wrote:

My guess is that FPGA PowerPC is going to be slower then Sam440 or Sam460,


The Vampire V4 is the fastest 68k variant available but much slower than a up to date 32Bit CPU should be, a FPGA PPC might reach the speed of old PowerUP cards.

Last edited by OneTimer1 on 29-Aug-2021 at 04:36 PM.

 Status: Offline
Profile     Report this post  
bison 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 29-Aug-2021 20:16:38
#7 ]
Super Member
Joined: 18-Dec-2007
Posts: 1874
From: N-Space

@Petah

More than three years, which is as close to "never" as the poll allows.

_________________
"Unix is supposed to fix that." -- Jay Miner

 Status: Offline
Profile     Report this post  
BigD 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 29-Aug-2021 21:22:48
#8 ]
Elite Member
Joined: 11-Aug-2005
Posts: 5910
From: UK

@Petah

Why are your polls so inane? Where’s the option for never?

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

 Status: Offline
Profile     Report this post  
agami 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 30-Aug-2021 6:42:12
#9 ]
Cult Member
Joined: 30-Jun-2008
Posts: 629
From: Melbourne, Australia

@BigD
Where’s the option for never?

@NutsAboutAmiga
My guess is that FPGA PowerPC is going to be slower then Sam440 or Sam460

@OneTimer1
When hell freezes over and the AOS4 sources are open sourced

@bison
... which is as close to "never" as the poll allows

Is there some inherent limit to the future scalability of FPGA that the rest of the industry is unaware of?
From what I'm reading it's clear that like many other technologies in the computing industry, FPGA will continue to scale in performance and in volume, and cost will continue to come down.
Furthermore, it's applicability in low volume production will increase.
Some predict a trend toward hybridization of CPU+FPGA, and even CPU+GPU+TPU+FPGA. None of which will eliminate the standalone FPGA implementations.

If you're talking about the small and anachronistic Amiga market, I'd rather spend $1,500 on an FPGA system with comparable performance of a SAM460, than $1,000 on an actual SAM460 (were there any). The latter is stuck being a SAM460, where the former can be changed.

 Status: Offline
Profile     Report this post  
Hammer 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 30-Aug-2021 9:08:57
#10 ]
Elite Member
Joined: 9-Mar-2003
Posts: 4270
From: Australia

@Petah

Altera FPGA is owned by Intel.
Xilinx FPGA is owned by AMD.

_________________
Core i9-9900K, DDR4-3800 32 GB RAM, GeForce RTX 3080 Ti
Ryzen 9 3900X, DDR4-3200 32 GB RAM, GeForce RTX 2080 Ti
Amiga 1200 (rev 1D1, KS 3.2, TF1260, 68060 @ 63 Mhz, 128 MB)
Amiga 500 (rev 6A, KS 3.2, 68K 50Mhz, 12 MB RAM)

 Status: Offline
Profile     Report this post  
BSzili 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 30-Aug-2021 16:15:15
#11 ]
Regular Member
Joined: 16-Nov-2013
Posts: 429
From: Unknown

@BigD

The best option would be pancakes!

_________________
This is just like television, only you can see much further.

 Status: Offline
Profile     Report this post  
IridiumFX 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 30-Aug-2021 17:02:35
#12 ]
Member
Joined: 7-Apr-2017
Posts: 57
From: London, UK

@agami

Yes, there is. The same old “clock speed” that screwed complex designs before.

I appreciate Achronix used to have 1.5GHz capable connection blocks in their designs, but you don’t implement an opcode in one logic element, hence you start halving the max speed for each step. Nowadays the same company advertises the speedster 7t at 750MHz capability, so don’t hold your breath.

Xilinx never never gunned for the GHz but offered slightly fatter logic elements. Altera used the other way around, with slightly faster interconnection and less capable LEs.

Just to say ... unless you plan to fill a virtex ultrascale+ of tiny RISC V cores, the complexity will bite your max clock speed and you’ll still hover at 80-150 MHz.
No matter how your FPGAs evolve, if you try to cram more complex designs into them, you’ll always be hitting the same wall.

Quote:

Is there some inherent limit to the future scalability of FPGA that the rest of the industry is unaware of?
From what I'm reading it's clear that like many other technologies in the computing industry, FPGA will continue to scale in performance and in volume, and cost will continue to come down.

 Status: Offline
Profile     Report this post  
AmigaBlitter 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 30-Aug-2021 17:52:39
#13 ]
Elite Member
Joined: 26-Sep-2005
Posts: 3499
From: Unknown

@IridiumFX

https://www.xilinx.com/products/silicon-devices/fpga/virtex-4q/modals/powerpc-32-bit-risc-processor.html

https://openpowerfoundation.org/cores-and-more/

https://dspace.mit.edu/handle/1721.1/46534

http://signal-processing.mil-embedded.com/articles/migrating-tools-renew-powerpc-applications/

Last edited by AmigaBlitter on 30-Aug-2021 at 05:56 PM.
Last edited by AmigaBlitter on 30-Aug-2021 at 05:54 PM.

_________________
retired

 Status: Offline
Profile     Report this post  
NutsAboutAmiga 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 30-Aug-2021 20:13:30
#14 ]
Elite Member
Joined: 9-Jun-2004
Posts: 11871
From: Norway

@AmigaBlitter

Quote:
Embedded 450 MHz, 700+ DMIPS RISC core (32-bit Harvard architecture).


Faster than BlizzardPPC, CyberStormPPC, slower then Sam440, in addition you need space for Amiga chipset or other hardware you’re simulating.

the question also becomes will cost more or less than a real PPC405 @ 450 Mhz, if real PPC405 is less expansive then it is pointless, unless thinking about putting custom stuff in it.

Last edited by NutsAboutAmiga on 30-Aug-2021 at 08:14 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

 Status: Offline
Profile     Report this post  
NutsAboutAmiga 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 30-Aug-2021 20:19:17
#15 ]
Elite Member
Joined: 9-Jun-2004
Posts: 11871
From: Norway

@BSzili

Yep, the pancakes option is best one, it’s not logical to do it, but this community is not wherry logical, so some who is a bit crazy might just do it, anyway, just because of challenge or something.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

 Status: Offline
Profile     Report this post  
IridiumFX 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 31-Aug-2021 0:18:33
#16 ]
Member
Joined: 7-Apr-2017
Posts: 57
From: London, UK

@AmigaBlitter
How’s that random collection of links related to my post?

The Virtex-4 had a real hard PPC core + fpga fabric.

Microwatt soft core runs at 100Mhz.

Chiselwatt is a port of the Microwatt core to chisel, so it is subject to the same speed constraints.

The other fancy paper just presents a research for a soft core ppc implementation and no stated speed grades.

Which point are you trying to make, my friend?

Quote:

AmigaBlitter wrote:
@IridiumFX

https://www.xilinx.com/products/silicon-devices/fpga/virtex-4q/modals/powerpc-32-bit-risc-processor.html

https://openpowerfoundation.org/cores-and-more/

https://dspace.mit.edu/handle/1721.1/46534

http://signal-processing.mil-embedded.com/articles/migrating-tools-renew-powerpc-applications/

 Status: Offline
Profile     Report this post  
matthey 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 31-Aug-2021 1:04:49
#17 ]
Super Member
Joined: 14-Mar-2007
Posts: 1145
From: Kansas

agami Quote:

Is there some inherent limit to the future scalability of FPGA that the rest of the industry is unaware of?
From what I'm reading it's clear that like many other technologies in the computing industry, FPGA will continue to scale in performance and in volume, and cost will continue to come down.
Furthermore, it's applicability in low volume production will increase.


The limit to FPGA scalability is the end of Moore's Law. FPGAs were playing catch up to modern chip fab processes as economies of scale allowed by building demand but they are now practically caught up so further improvements likely won't seam as fast. The cheapest FPGAs likely won't get much cheaper while more powerful FPGAs will continue to improve value (price, size, features) but with diminishing improvements. FPGAs face the same chip fab wall and current leakage problems as ASICs which will make chip die shrinks more and more expensive and slower to develop. Large FPGA designs are more likely to suffer from routing inefficiencies than ASIC designs.

IridiumFX Quote:

Yes, there is. The same old “clock speed” that screwed complex designs before.


Complex core designs offer better performance/MHz and performance/Watt. High clocked simple RISC cores were a failure. FPGAs do suffer more from routing inefficiencies than ASICs which make high performance FPGA core "complex designs" more challenging.

IridiumFX Quote:

I appreciate Achronix used to have 1.5GHz capable connection blocks in their designs, but you don’t implement an opcode in one logic element, hence you start halving the max speed for each step. Nowadays the same company advertises the speedster 7t at 750MHz capability, so don’t hold your breath.


The worst case routing of larger FPGAs at smaller die sizes grows. Clock skew can be a problem for small chip die sizes with ASIC designs too. Transistors and connections are locally faster but accessing resources far away is difficult because electricity takes longer to travel there and larger connections are needed. The optimum core design in FPGA may be simpler than in an ASIC but an ASIC allows for a more "complex design" due to efficient placement of resources and routing that FPGA cores can't compete with.

IridiumFX Quote:

Just to say ... unless you plan to fill a virtex ultrascale+ of tiny RISC V cores, the complexity will bite your max clock speed and you’ll still hover at 80-150 MHz.
No matter how your FPGAs evolve, if you try to cram more complex designs into them, you’ll always be hitting the same wall.


Simple multi-core designs have one advantage. The clocks of mostly standby cores can be reduced to save power but FPGA cores use more electricity than an ASIC so there is no technical advantage for simple FPGA cores. FPGA cores are usually 1-2 simple cores in cheap and small FPGAs for low volume production where price is more important than power reduction or performance. I don't expect this to change much.

NutsAboutAmiga Quote:

Faster than BlizzardPPC, CyberStormPPC, slower then Sam440, in addition you need space for Amiga chipset or other hardware you’re simulating.


The CyberstormPPC 604e is a significantly more powerful core than the weak Sam440 core. A 200-400MHz 604e can likely hold its own in benchmarks against the weak PPC405@450MHz design despite the much older chip fab process and memory technology.

NutsAboutAmiga Quote:

the question also becomes will cost more or less than a real PPC405 @ 450 Mhz, if real PPC405 is less expansive then it is pointless, unless thinking about putting custom stuff in it.


A CPU+FPGA is unlikely to be cheaper than a CPU only. Customization for specialized applications is more interesting although RISC-V may be a better candidate not because the core would be simpler but because RISC-V has much more free encoding space for custom FPGA instructions. There are open hardware RISC-V cores so maybe it is just a matter of time to replace the PPC core.

 Status: Offline
Profile     Report this post  
fishy_fis 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 31-Aug-2021 3:03:55
#18 ]
Super Member
Joined: 29-Mar-2004
Posts: 1942
From: Australia

@Petah

Never.
It won't reach a point it makes economical sense for probably a decade and ppc Amigas are a niche inside a niche inside a niche.
Its not like 68k were there was millions upon millions of loyal enthusiasts.
With PPC theres a few thousand users at best, and judging by forums comments apparently half of them aren't exactly happy with the ppc era either.

How is all this not painfully obvious?
Or is the Amiga just so dead that flights of fantasy on forums is all Amiga fans have left now?

 Status: Offline
Profile     Report this post  
NutsAboutAmiga 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 31-Aug-2021 8:11:42
#19 ]
Elite Member
Joined: 9-Jun-2004
Posts: 11871
From: Norway

@fishy_fis

Too many gamers, too few content creators, its kind a sign, when see new look alike amiga plastic keyboard that don’t work, created purely for esthetics, console form factor like Amiga systems, as home computer the Amiga is struggling to stay relevant.

This way we need web browsers, and music players, media players, and modern programs, for productivity, to get people keep using it, it has to be convenient.

The adaptation of PowerPC should happen back in 1995 by ESCOM, when community was pretty big, A1200 with PowerPC as standard, no 680x0, just a JIT compiler, and AmigaOS3.3 PowerPC, similar to MacOS7.x.x approach apple did. Plus, something like SAGA chipset, as a stepping stone, into true colors.

Personally loved if A1200 was mode modular computer, easier to upgrade, more adapted to times, A1200 without CDROM, and without HD Floppy drives was wrong product back in 1992.

Last edited by NutsAboutAmiga on 31-Aug-2021 at 10:41 AM.
Last edited by NutsAboutAmiga on 31-Aug-2021 at 08:40 AM.
Last edited by NutsAboutAmiga on 31-Aug-2021 at 08:39 AM.
Last edited by NutsAboutAmiga on 31-Aug-2021 at 08:34 AM.
Last edited by NutsAboutAmiga on 31-Aug-2021 at 08:20 AM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

 Status: Offline
Profile     Report this post  
AmigaBlitter 
Re: [POLL] Brand new FPGA-based PPC cards - when will they hit the market?
Posted on 31-Aug-2021 8:18:04
#20 ]
Elite Member
Joined: 26-Sep-2005
Posts: 3499
From: Unknown

@IridiumFX



Nothing in particular, just a few information page got here and there.

_________________
retired

 Status: Offline
Profile     Report this post  
Goto page ( 1 | 2 Next Page )

[ home ][ about us ][ privacy ] [ forums ][ classifieds ] [ links ][ news archive ] [ link to us ][ user account ]
Copyright (C) 2000 - 2019 Amigaworld.net.
Amigaworld.net was originally founded by David Doyle