Click Here
home features news forums classifieds faqs links search
6071 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
16 crawler(s) on-line.
 130 guest(s) on-line.
 0 member(s) on-line.



You are an anonymous user.
Register Now!
 A1200:  7 mins ago
 retrofaza:  12 mins ago
 pixie:  18 mins ago
 OlafS25:  31 mins ago
 BigD:  32 mins ago
 zErec:  39 mins ago
 amigakit:  42 mins ago
 kolla:  1 hr 25 mins ago
 edwardsjethro:  2 hrs 17 mins ago
 joeyunderwood:  2 hrs 18 mins ago

/  Forum Index
   /  Amiga General Chat
      /  Trevor Dickinson nominated for comment of the year! :-)
Register To Post

Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 Next Page )
PosterThread
MEGA_RJ_MICAL 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 0:30:12
#181 ]
Super Member
Joined: 13-Dec-2019
Posts: 1200
From: AMIGAWORLD.NET WAS ORIGINALLY FOUNDED BY DAVID DOYLE

Ok but what is better,
packed or planar?

_________________
I HAVE ABS OF STEEL
--
CAN YOU SEE ME? CAN YOU HEAR ME? OK FOR WORK

 Status: Offline
Profile     Report this post  
agami 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 2:32:52
#182 ]
Super Member
Joined: 30-Jun-2008
Posts: 1637
From: Melbourne, Australia

@OlafS25

Quote:
4 Million? Interesting. I also read somewhere in a interview that one of Hyperion claimed that they invested somewhere around 2.5 Million dollar in amigaos (including lawsuits of course). So asking 4 million sounds not unrealistic from my point of view.

$4M is what I have extracted based on my interactions. Ben claims $10M.

Also, I am not including legal costs as Ben was doing most of the work, despite how it might have been put down on the books.


Quote:
But that is not how the value of a product is defined. Nobody is interested what you "might" have invested in the past but what you can earn with it in future, And of course what investments are needed to get it in a state to sell it again.

Correct. That is the essence of capitalism. When it comes to selling company/IP it comes down to what it is worth to the buyer.

But when we are discussing "profitability", then we use P&L statements.
Even at my conservative assessment of $4M, AmigaOS 4 is not profitable. At Ben's $10M, it will never be. His only hope, and that of his investors, is to sell off the assets for $10M+.

Mind you, the investors would have already written off the loss over the years, and if they know how to do it right then they've already recovered their losses in tax savings. So even selling the assets at $10M, everyone essentially doubles their money (pre-tax).

_________________
All the way, with 68k

 Status: Offline
Profile     Report this post  
matthey 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 4:12:33
#183 ]
Super Member
Joined: 14-Mar-2007
Posts: 1968
From: Kansas

BigD Quote:

OK. maybe a custom ASIC could have brought costs down but no one with the skills was available. AmiBerry IS a great solution and it's sad the coder will never get rich off from the work he did. Yes, the implementation from RGL isn't ideal but it's functional and Pandory/AMiNIMiga take up the slack IMHO. I was playing The Settlers tonight and although there is a bug on Pandory V2 at present with the game (I overcome it by starting it by loading a Save State) it is still great emulation and with 2-player mouse support. Deluxe Paint V works great on AMiNIMiga V200. I own real 68k Amigas but for 99% of people who don't need serial or parallel functionality, I'd recommend THEA500 Mini.


There are at least 4 AGA compatible Amiga chipset cores that have been created in HDL. These can be used in an FPGA or an ASIC created from them. Each was mostly programmed by a single person. Most have enhancements like digital video output (HDMI/DVI), chunky/RTG support, 16 bit audio and more modern memory support (they aren't slow like AGA). This list does not include Jeri's chipset recreation or other OCS/ECS Amiga chipsets. The Amiga is blessed to have many HDL programmers that can do this work. More advanced verification and testing would need to be done for an ASIC but a big part of the core design work is already done. Like I said, the chipset is not the problem. An advanced competitive 68k CPU core is more difficult to develop but also where much of the competitiveness of a SoC would come from. We need better 68k hardware than a MiniMig. There is plenty of low end hardware already.

Hammer Quote:

FYI, as of April 2022, TheA500Mini has sold 186,000 units.


No. It was the "C64 direct to TV joystick" that sold 186,000 units in the first hour. THEA500Mini wasn't cost reduced enough and lacks value to sell that well as a toy. Maybe if they had gone the ASIC route it would have been possible but they are a tiny business and the Amiga is more advanced to develop. The hardware production cost could approach Raspberry Pi product territory if mass produced even with 1-2GHz 68060 CPU performance, AA+ chipset and maybe even an Imagination Technologies hybrid raytracing 3D GPU. Development and licensing costs are a bigger cost concern than the area/transistors making the SoC ASIC too expensive to mass produce.

Hammer Quote:

ARM Cortex A53 has the performance of Cortex A9 with 64-bit and SIMD NEON.


The ARM Cortex-A9 is the older but popular 32 bit OoO core. The integer performance is better than the in-order ARM Cortex-A53 although the standard AArch64 features likely make it better performance for some workloads that can take advantage of 64 bit and SIMD use.

DMIPS/MHz/core
1.56 68060 (32 bit in-order)
1.83 ColdFire V5 (32 bit in-order)
1.88 Pentium P54C (32 bit in-order)
1.90 Cortex-A7 (32 bit in-order)
2.03 Raspberry Pi 3 Cortex-A53 (64 bit in-order)
2.40 A1222 P1022 (32 bit OoO)
2.50 Cortex-A9 (32 bit OoO)
3.76 Raspberry Pi 4 Cortex-A72 (64 bit OoO)

A 64 bit CPU needs more resources than a 32 bit CPU. The 64 bit Cortex-A53 struggles with the minimalist resources and increased code size of 64 bit AArch64. A good example is the 7-Zip benchmark where Thumb2 has better single core compression performance and the original ARM32 has better single core decompression performance than AArch64 mode on the Cortex-A53.

https://www.7-cpu.com/

Judging by how much the Cortex-A53 struggles with fat AArch64, this core looks near the limit of how far AArch64 scales down which isn't that far. ARM has Thumb2 to cover the low end and 32 bit market though. AArch64 code density is unimpressive but there isn't much competition for 64 bit ISAs, primarily x86-64 and RISC-V.

The DMIPS/MHz/core numbers above are a rough measure of integer performance. The 68060 number is official from Motorola but they were conservative in estimates and lacked compiler support for the 68060. Some integer benchmarks have the 68060 ahead of the Pentium which is already competing with the ARM in-order CPUs. This is impressive considering the age of the 68060 and Pentium process and 1/4 of the L1 caches.

 Status: Offline
Profile     Report this post  
cdimauro 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 4:49:16
#184 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3621
From: Germany

@matthey

Quote:

matthey wrote:
Do we understand what we miss(ed)? Actually, we were likely close to viable ASIC Amiga recreation hardware on other occasions including the BoXeR and Natami projects. There are multiple advanced FPGA Amiga chipset recreations that could be used. The tough part is the CPU and the road blocks. Some people would rather have their hobby PPC hardware recreations using PC hardware (Trevor), emulation (half the Amiga community that thinks an ASIC would cost too much) or a pile of cash and complete control of everything (Ben) instead. The Amiga is running out of opportunities. Another lost decade and the Amiga legacy is sealed.

I fully understand Matt, but what's the problem if some people want to purse OTHER ways?

If they like PowerPCs then let them continue with those.

If they like emulation then let them continue with it.

You can find someone which likes 68ks and you go with that.

Choices aren't mutually-exclusive.
Quote:

matthey wrote:

There are at least 4 AGA compatible Amiga chipset cores that have been created in HDL. These can be used in an FPGA or an ASIC created from them. Each was mostly programmed by a single person. Most have enhancements like digital video output (HDMI/DVI), chunky/RTG support, 16 bit audio and more modern memory support (they aren't slow like AGA). This list does not include Jeri's chipset recreation or other OCS/ECS Amiga chipsets. The Amiga is blessed to have many HDL programmers that can do this work. More advanced verification and testing would need to be done for an ASIC but a big part of the core design work is already done. Like I said, the chipset is not the problem.

I think the opposite: it's The problem because all current implementations aren't 100% accurate.

And it's a double problem because the Amiga Mini (let's call it this way) market is 100% for games.

However games require 100% compatibility and that's why the problem is repeated twice: you need a 68000+OCS/ECS compatible system and a 68EC020+AGA compatibile system, both on the same ASIC and which can be selected at boot time.

This is the only solution that allows you to tackle the (full) compatibility which is request by the games.
Quote:
An advanced competitive 68k CPU core is more difficult to develop but also where much of the competitiveness of a SoC would come from. We need better 68k hardware than a MiniMig. There is plenty of low end hardware already.

You don't need it for games. A 7Mhz 68000 and a 14Mhz 68EC020 core are both enough AND required if you want to support in the best way the Amiga game market.

Which is the Amiga Mini market.

The only market, I would say, because I doubt that people that bought the C64 Mini before and the Amiga Mini after for running applications...
Quote:
The ARM Cortex-A9 is the older but popular 32 bit OoO core. The integer performance is better than the in-order ARM Cortex-A53 although the standard AArch64 features likely make it better performance for some workloads that can take advantage of 64 bit and SIMD use.

DMIPS/MHz/core
1.56 68060 (32 bit in-order)
1.83 ColdFire V5 (32 bit in-order)
1.88 Pentium P54C (32 bit in-order)
1.90 Cortex-A7 (32 bit in-order)
2.03 Raspberry Pi 3 Cortex-A53 (64 bit in-order)
2.40 A1222 P1022 (32 bit OoO)
2.50 Cortex-A9 (32 bit OoO)
3.76 Raspberry Pi 4 Cortex-A72 (64 bit OoO)

A 64 bit CPU needs more resources than a 32 bit CPU. The 64 bit Cortex-A53 struggles with the minimalist resources and increased code size of 64 bit AArch64. A good example is the 7-Zip benchmark where Thumb2 has better single core compression performance and the original ARM32 has better single core decompression performance than AArch64 mode on the Cortex-A53.

https://www.7-cpu.com/

Judging by how much the Cortex-A53 struggles with fat AArch64, this core looks near the limit of how far AArch64 scales down which isn't that far. ARM has Thumb2 to cover the low end and 32 bit market though.

Exactly. And both cores are here to stay for ARM: there's no other solution and I don't think that I'll come in future. ARM has set its stones...
Quote:
AArch64 code density is unimpressive but there isn't much competition for 64 bit ISAs, primarily x86-64 and RISC-V.

Good for competitors...
Quote:
The DMIPS/MHz/core numbers above are a rough measure of integer performance. The 68060 number is official from Motorola but they were conservative in estimates and lacked compiler support for the 68060. Some integer benchmarks have the 68060 ahead of the Pentium which is already competing with the ARM in-order CPUs. This is impressive considering the age of the 68060 and Pentium process and 1/4 of the L1 caches.

Indeed. The weaker are the cores the more CISC microarchitectures perform compared to equivalent RISC ones.


@MEGA_RJ_MICAL

Quote:

MEGA_RJ_MICAL wrote:
Ok but what is better,
packed or planar?

Ehm.

 Status: Offline
Profile     Report this post  
matthey 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 4:50:07
#185 ]
Super Member
Joined: 14-Mar-2007
Posts: 1968
From: Kansas

agami Quote:

$4M is what I have extracted based on my interactions. Ben claims $10M.

Also, I am not including legal costs as Ben was doing most of the work, despite how it might have been put down on the books.


How much should Hyperion's AmigaOS 4 be discounted considering they didn't deliver the source after it was payed for last time? Does the asking price exclude the part of AmigaOS 4 that is owned by others? I wonder if Amiga Inc. (Delaware) included AmigaOS 4 they owned in their $87 million business valuation. Amiga Corporation should own more AmigaOS assets than Hyperion and would be safer to deal with. There are other choices to place on top of AmigaOS 3.1 like MorphOS, AROS or the Enhancer software even if Hyperion wants to take the parts of AmigaOS 4 they may own to the grave. The lawyer in charge of their bankruptcy should be much more reasonable to deal with.

 Status: Offline
Profile     Report this post  
kolla 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 5:54:42
#186 ]
Elite Member
Joined: 20-Aug-2003
Posts: 2859
From: Trondheim, Norway

@matthey

Those references are not about “the original a500 mini plan” those are anecdotes about exploring the possibilities of FPGA 15-20 years ago. Minimig fulfilled that idea.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

 Status: Offline
Profile     Report this post  
matthey 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 6:18:54
#187 ]
Super Member
Joined: 14-Mar-2007
Posts: 1968
From: Kansas

cdimauro Quote:

I fully understand Matt, but what's the problem if some people want to purse OTHER ways?

If they like PowerPCs then let them continue with those.

If they like emulation then let them continue with it.

You can find someone which likes 68ks and you go with that.

Choices aren't mutually-exclusive.


Normally, I would say that is ok. The problem is that watching the PPC money go down the drain for a few hundred existing users and emulation for a few thousand existing users won't do anything for the Amiga. Neither bring in new and younger Amiga users. Competitive mass produced low end hardware could bring in new younger users like we often started using cheap low end computers when we were young. The younger generations are more likely to know what a Raspberry Pi is today than an Amiga. The Amiga is so much more fun and has retro games which is more appealing than the Raspberry Pi. Rather than pool resources and cooperate to produce competitive low end hardware, we watch the Amiga legacy go down the drain too.

cdimauro Quote:

I think the opposite: it's The problem because all current implementations aren't 100% accurate.

And it's a double problem because the Amiga Mini (let's call it this way) market is 100% for games.

However games require 100% compatibility and that's why the problem is repeated twice: you need a 68000+OCS/ECS compatible system and a 68EC020+AGA compatibile system, both on the same ASIC and which can be selected at boot time.

This is the only solution that allows you to tackle the (full) compatibility which is request by the games.


100% compatibility is difficult. Even the TG68 in FPGA or ASIC is not 100% compatible with the 68000 or 68020 because it is not cycle exact. This is rarely a problem for the Amiga but it can be a problem for some 68k hardware like the Atari ST. Either eFPGA blocks in the ASIC or a small separate FPGA for low end CPUs and chipsets would be very versatile. A $10-$15 FPGA would likely be enough. It could even have its own HDMI output.

cdimauro Quote:

You don't need it for games. A 7Mhz 68000 and a 14Mhz 68EC020 core are both enough AND required if you want to support in the best way the Amiga game market.

Which is the Amiga Mini market.

The only market, I would say, because I doubt that people that bought the C64 Mini before and the Amiga Mini after for running applications...


The idea is to target retro gaming but make it a full and powerful general purpose computer too. The retro games and low price hooks customers but the AmigaOS would be there in the background with an easy to use GUI and I/O resembling a PC or at least RPi. The RPi is cheap and general purpose but doesn't have much retro appeal or FPGA features. MiSTer has a big FPGA and is easy to use but is expensive, runs hot and isn't a general purpose computer. The RPi uses smart phone SoCs and MiSTer uses a FPGA developer board. Can purpose built mass produced hardware do better?

kolla Quote:

Those references are not about “the original a500 mini plan” those are anecdotes about exploring the possibilities of FPGA 15-20 years ago. Minimig fulfilled that idea.


"Exploring the possibilities of FPGA" is not what it sounds like to me when Jeri talks about "making the full custom ASIC".

Last edited by matthey on 29-Oct-2022 at 06:24 AM.

 Status: Offline
Profile     Report this post  
NutsAboutAmiga 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 11:52:46
#188 ]
Elite Member
Joined: 9-Jun-2004
Posts: 12795
From: Norway

@matthey

Quote:
The problem is that watching the PPC money go down the drain


Where do you think that monay comes from? and why is wrong pay for something we like?
its not your monay why is your problem?

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

 Status: Offline
Profile     Report this post  
NutsAboutAmiga 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 12:01:22
#189 ]
Elite Member
Joined: 9-Jun-2004
Posts: 12795
From: Norway

@Hammer

Quote:

TheA500Mini can support AGA, virtual 68040 CPU can up to 268 MIPS and 187 MFLOPS backed by quad-core ARM Cortex A53 @ 1.8 GHz (ref 2).

TheA500Mini has the potential surplus A600 and A1200 install base.

TheA500Mini's GPU is ARM Mali-T720 MP2 complies with OpenGL ES 3.1, OpenCL 1.1, and DirectX11 feature set hardware.

TheA500Mini's 186,000 units sale blows away PowerPC neo-Amigas and Apollo-Core's Vampires.

TheA500Mini substantially increased AGA and super-68040 with up to 268 MIPS Amiga userbase.


it sounds like you never used one..
you don’t have a Picasso96 drivers for “ARM Mali-T720 MP2” that truly supports.
there is not WARP 3D drivers with OpenGL ES support on Classic AmigaOS.
why do you even talk about DirectX11???? Are running windows on it, or something else?

What you have is SW emulation trying to emulate OCS/AGA, and a frame buffer for P96 on top of a host OS, that’s far, far away from a ideal.

When I tried the TheA500 mini, I did not think it played that well, I’m sure there better FPGA systems that do better job, at running classic games with less latency and lag. I’m pretty sure my PC can do better. So I don’t get the point of that thing, besides making money from fools who don’t know better. If WinUAE is faster, works better and is free, why spend that money. Crazy.

Its basally a raspberry pi in some plastic, with none working keyboard, without explanation slots. Without any upgrade possibilities at all. to me its not worth the monay.

Last edited by NutsAboutAmiga on 29-Oct-2022 at 12:23 PM.
Last edited by NutsAboutAmiga on 29-Oct-2022 at 12:04 PM.
Last edited by NutsAboutAmiga on 29-Oct-2022 at 12:02 PM.

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

 Status: Offline
Profile     Report this post  
fishy_fis 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 15:54:26
#190 ]
Elite Member
Joined: 29-Mar-2004
Posts: 2156
From: Australia

@NutsAboutAmiga

err.......

Quote:
and why is wrong pay for something we like?
its not your monay why is your problem?


And then it your very next comment:

Quote:
So I don’t get the point of that thing, besides making money from fools who don’t know better


So what you're suggesting is it's only foolish if *YOU* personally have no interest then?
Interesting.

Also, I didnt realise WinUAE came with licensed kickstart rom images or 24 licensed games, a mouse and a joypad, nor an ARM SoC included in the price of free.

Yes, WinUAE is faster. I get about 5500 mips of emulated 68k performance on my Ryzen 5900x and about 7500 mips of emulated 68k performance on my new i9-13900k, but I still bought an a500 mini.
Fun little machine and I knew exactly what I was buying; licensed software, and a fun little machine using immeasurably less power than my x86 boxes that can run 68k software at ~%500 the speed of a stock '060.

"its not your monay why is your problem?"

 Status: Offline
Profile     Report this post  
kolla 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 19:13:02
#191 ]
Elite Member
Joined: 20-Aug-2003
Posts: 2859
From: Trondheim, Norway

@matthey

Talking about something isn’t the same as having a plan. I mean, Gunnar has been talking about ASIC since forever. And you yourself talk about 68k ASIC pretty much constantly, but there’s hardly a plan.

Btw - from where do you have it that MiSTer “runs hot”? Even overclocked to 1.2 GHz it barely gets to human fever temperatures. Also, it _can_ be used as a general purpose system, I installed Gentoo on mine to have more software.

Last edited by kolla on 29-Oct-2022 at 07:20 PM.
Last edited by kolla on 29-Oct-2022 at 07:17 PM.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

 Status: Offline
Profile     Report this post  
NutsAboutAmiga 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 20:12:53
#192 ]
Elite Member
Joined: 9-Jun-2004
Posts: 12795
From: Norway

@fishy_fis

Your taking things out of context, Hammer compared it to a PowerPC / AmigaOS4.x system.
It's not my fault.

"its not your monay why is your problem?"

Well its not, your free to buy it, if want do that, its just nothing for me.

Last edited by NutsAboutAmiga on 29-Oct-2022 at 08:14 PM.
Last edited by NutsAboutAmiga on 29-Oct-2022 at 08:14 PM.
Last edited by NutsAboutAmiga on 29-Oct-2022 at 08:13 PM.

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

 Status: Offline
Profile     Report this post  
cdimauro 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 20:14:14
#193 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3621
From: Germany

@matthey

Quote:

matthey wrote:
cdimauro Quote:

I fully understand Matt, but what's the problem if some people want to purse OTHER ways?

If they like PowerPCs then let them continue with those.

If they like emulation then let them continue with it.

You can find someone which likes 68ks and you go with that.

Choices aren't mutually-exclusive.


Normally, I would say that is ok. The problem is that watching the PPC money go down the drain for a few hundred existing users and emulation for a few thousand existing users won't do anything for the Amiga.

I know, but we're on the internet age and people are very well informed. I don't think that someone changes its mind at this point in time.
Quote:
Neither bring in new and younger Amiga users.

Why not? Youngs are used to play games on any possible device.
Quote:
Competitive mass produced low end hardware could bring in new younger users like we often started using cheap low end computers when we were young. The younger generations are more likely to know what a Raspberry Pi is today than an Amiga. The Amiga is so much more fun and has retro games which is more appealing than the Raspberry Pi. Rather than pool resources and cooperate to produce competitive low end hardware, we watch the Amiga legacy go down the drain too.

Then you've to provide a very competitive product which should be more attractive compared to the RPi.
Quote:
cdimauro Quote:

I think the opposite: it's The problem because all current implementations aren't 100% accurate.

And it's a double problem because the Amiga Mini (let's call it this way) market is 100% for games.

However games require 100% compatibility and that's why the problem is repeated twice: you need a 68000+OCS/ECS compatible system and a 68EC020+AGA compatibile system, both on the same ASIC and which can be selected at boot time.

This is the only solution that allows you to tackle the (full) compatibility which is request by the games.


100% compatibility is difficult.

Indeeed. That's why WHDLoad gives a big help here.
Quote:
Even the TG68 in FPGA or ASIC is not 100% compatible with the 68000 or 68020 because it is not cycle exact.

For the 68020 isn't that much important, because at the time (Amiga 1200 & 4000) coders were a bit more aware of processors differences.

But for the 68000 it can be very important, unfortunately. It's strange that there's no cycle-exact 68000. AFAIR the core in WinUAE does it, so that it should be enough to copy the same implementation.
Quote:
This is rarely a problem for the Amiga but it can be a problem for some 68k hardware like the Atari ST. Either eFPGA blocks in the ASIC or a small separate FPGA for low end CPUs and chipsets would be very versatile.

That's exactly what I've suggested. At the very end a 68000 + OCS/ECS shouldn't take so much space / LEs.
Quote:
A $10-$15 FPGA would likely be enough. It could even have its own HDMI output.

It should have it.
Quote:
cdimauro Quote:

You don't need it for games. A 7Mhz 68000 and a 14Mhz 68EC020 core are both enough AND required if you want to support in the best way the Amiga game market.

Which is the Amiga Mini market.

The only market, I would say, because I doubt that people that bought the C64 Mini before and the Amiga Mini after for running applications...

The idea is to target retro gaming but make it a full and powerful general purpose computer too. The retro games and low price hooks customers but the AmigaOS would be there in the background with an easy to use GUI and I/O resembling a PC or at least RPi.

Then you could have a solution with 3 implementations embedded: 7Mhz 68000 + OCS/ECS, 14Mhz 68EC020 + AGA, 68020+ at maximum speed + AGA. Maybe AGA could be easily shared by the last two implementations.
Quote:
The RPi is cheap and general purpose but doesn't have much retro appeal or FPGA features. MiSTer has a big FPGA and is easy to use but is expensive, runs hot and isn't a general purpose computer. The RPi uses smart phone SoCs and MiSTer uses a FPGA developer board. Can purpose built mass produced hardware do better?

It depends on how much it costs. I mean: the overall TCO.

 Status: Offline
Profile     Report this post  
matthey 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 29-Oct-2022 23:02:43
#194 ]
Super Member
Joined: 14-Mar-2007
Posts: 1968
From: Kansas

kolla Quote:

Talking about something isn’t the same as having a plan. I mean, Gunnar has been talking about ASIC since forever. And you yourself talk about 68k ASIC pretty much constantly, but there’s hardly a plan.


There is too much uncertainty to develop a bigger plan. Uncertainty is the killer of business plans. It makes risk assessment too hard. The same is true for Amiga Corporation. Some people think Amiga Corporation should do more which they could but they are being conservative and maintaining financial discipline. A good example is when Michele Battilana was asked about an Amiga Maxi. They may have discussed plans with business partners but there is too much uncertainty to move forward despite opportunities for the Amiga. In contrast, Hyperion has gambled and gambled on their gambles. Go to the casino, keep betting everything and it won't take long to lose everything.

kolla Quote:

Btw - from where do you have it that MiSTer “runs hot”? Even overclocked to 1.2 GHz it barely gets to human fever temperatures. Also, it _can_ be used as a general purpose system, I installed Gentoo on mine to have more software.


I have read on forums that the FPGA and some other chips run hot, at least with the modular add-on boards added which is common. There may be problems that are a combo of heat and lack of power. The add-on memory doesn't work at higher clock speeds for some people and HDMI with anything but super short high quality cables flakes out for others (some Vampire hardware had similar problems). The modular connections and electrical distances of add-on boards may cause problems as well. A bigger and better ground can solve some issues but there are already grounds in the connections. As much as the modularity is flexible, it would be better to have a single SBC that has everything needed.

cdimauro Quote:

Why not? Youngs are used to play games on any possible device.


With PPC hardware, there is no hardware close to being competitive in performance/price while younger people often have less money and are thrifty. With emulation, there is a leaning curve and there may be a lack of ROMs while the gaming quality can be low due to lag/latency. THEA500 Mini did a nice job of presentation with minimal out of the box configuration so it is possible to do better. It should already be possible to distribute an emulator+ROMs with licensing for an auto running Amiga program in a 68k Amiga VM. It's never going to be as attractive as native competitive hardware though.

cdimauro Quote:

For the 68020 isn't that much important, because at the time (Amiga 1200 & 4000) coders were a bit more aware of processors differences.

But for the 68000 it can be very important, unfortunately. It's strange that there's no cycle-exact 68000. AFAIR the core in WinUAE does it, so that it should be enough to copy the same implementation.


I believe it would be possible to make a cycle exact 68000 in FPGA although it would be much more work. It is surprising it hasn't been done by Atari fans but this likely an indication of how compatible the TG68 is without cycle exact timing. I have doubts about how cycle exact emulation can be. It is difficult for software code to get down to the granularity of a single cycle.

cdimauro Quote:

Then you could have a solution with 3 implementations embedded: 7Mhz 68000 + OCS/ECS, 14Mhz 68EC020 + AGA, 68020+ at maximum speed + AGA. Maybe AGA could be easily shared by the last two implementations.


Right, and these are probably best provided by FPGA functionality which can simulate chipsets and low performance CPUs for other computers and consoles too. It could even be used for better than software acceleration that requires parallel, DSP or specialized processing. FPGA technology adds a lot of versatility and is cheap at the low end.

Last edited by matthey on 29-Oct-2022 at 11:06 PM.
Last edited by matthey on 29-Oct-2022 at 11:05 PM.

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 30-Oct-2022 5:27:52
#195 ]
Elite Member
Joined: 9-Mar-2003
Posts: 5246
From: Australia

@NutsAboutAmiga

Quote:

it sounds like you never used one..
you don’t have a Picasso96 drivers for “ARM Mali-T720 MP2” that truly supports.

I didn't claim AmigaOS side supports ARM Mali-T720 MP2. I merely stated ARM Mali-T720 MP2 hardware availability with the device.

Quote:

why do you even talk about DirectX11???? Are running windows on it, or something else?

To convey the GPU's feature set in PC GPU terms.

https://www.cnx-software.com/2017/04/01/allwinner-h6-processor-for-4k-hdr-set-top-boxes-supports-usb-3-0-pcie-and-smart-card-interfaces/
3D GPU – Dual shader ARM Mali-T720 with support for OpenGL ES3.1/3.0/2.0/1.1, OpenCL 1.1/RenderScript, Microsoft DirectX 11 FL9_3

Direct3D 9/10/11 can operate from DXVK without actual Microsoft Windows.

Quote:

What you have is SW emulation trying to emulate OCS/AGA, and a frame buffer for P96 on top of a host OS, that’s far, far away from a ideal.

Name an alternative product with a similar TheA500 mini's price range and included case design.

Quote:

When I tried the TheA500 mini, I did not think it played that well, I’m sure there better FPGA systems that do better job, at running classic games with less latency and lag.

You are welcome to produce alternative FPGA-based A500 products with the TheA500 mini's price.

https://www.minimig.ca/index.php/product/minimig11/
FPGA-based Minimig v1.1 is priced at $220.

The A500 Mini has $125 price range.

Quote:

Its basally a raspberry pi in some plastic, with none working keyboard, without explanation slots.

Amiga CD32 doesn't have a keyboard. A500 Mini has 3 x USB 2.0 ports and USB 2.0 port can reach 53 MByte/s.

Amiga CD32/A1200's expansion bus is slower than USB 2.0.

Last edited by Hammer on 30-Oct-2022 at 05:28 AM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 30-Oct-2022 5:35:18
#196 ]
Elite Member
Joined: 9-Mar-2003
Posts: 5246
From: Australia

@matthey

Quote:
The ARM Cortex-A9 is the older but popular 32 bit OoO core. The integer performance is better than the in-order ARM Cortex-A53 although the standard AArch64 features likely make it better performance for some workloads that can take advantage of 64 bit and SIMD use.

DMIPS/MHz/core
1.56 68060 (32 bit in-order)
1.83 ColdFire V5 (32 bit in-order)
1.88 Pentium P54C (32 bit in-order)
1.90 Cortex-A7 (32 bit in-order)
2.03 Raspberry Pi 3 Cortex-A53 (64 bit in-order)
2.40 A1222 P1022 (32 bit OoO)
2.50 Cortex-A9 (32 bit OoO)
3.76 Raspberry Pi 4 Cortex-A72 (64 bit OoO)

A 64 bit CPU needs more resources than a 32 bit CPU. The 64 bit Cortex-A53 struggles with the minimalist resources and increased code size of 64 bit AArch64. A good example is the 7-Zip benchmark where Thumb2 has better single core compression performance and the original ARM32 has better single core decompression performance than AArch64 mode on the Cortex-A53.

https://www.7-cpu.com/

Judging by how much the Cortex-A53 struggles with fat AArch64, this core looks near the limit of how far AArch64 scales down which isn't that far. ARM has Thumb2 to cover the low end and 32 bit market though. AArch64 code density is unimpressive but there isn't much competition for 64 bit ISAs, primarily x86-64 and RISC-V.

The DMIPS/MHz/core numbers above are a rough measure of integer performance. The 68060 number is official from Motorola but they were conservative in estimates and lacked compiler support for the 68060. Some integer benchmarks have the 68060 ahead of the Pentium which is already competing with the ARM in-order CPUs. This is impressive considering the age of the 68060 and Pentium process and 1/4 of the L1 caches.


From https://community.arm.com/arm-community-blogs/b/architectures-and-processors-blog/posts/the-top-5-things-to-know-about-cortex-a53



From ARM's claims
At the same frequency, Cortex-A53 delivers more than 20% higher instruction throughput than the Cortex-A9 for representative workloads.

the Cortex-A53 achieves higher single-thread performance by pushing a simpler design farther – some of the key factors enabling the performance of the Cortex-A53 include the integrated low latency level 2 cache, the larger 512 entry main TLB, and the complex branch predictor


Last edited by Hammer on 30-Oct-2022 at 05:39 AM.
Last edited by Hammer on 30-Oct-2022 at 05:37 AM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 30-Oct-2022 5:52:47
#197 ]
Elite Member
Joined: 9-Mar-2003
Posts: 5246
From: Australia

@NutsAboutAmiga

Quote:
Your taking things out of context, Hammer compared it to a PowerPC / AmigaOS4.x system.


If I'm going to compare with PowerPC / AmigaOS4.x system against something, I'll do it with my Windows 11 gaming PC and will crush it.

My budget for PiStorm with RPI 3a+ is throw-away money that doesn't affect my budget for my next gaming PC.

Wake me up when A1220 reaches Vampire's 10,000 units.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 30-Oct-2022 6:09:15
#198 ]
Elite Member
Joined: 9-Mar-2003
Posts: 5246
From: Australia

@Rob

Quote:

Rob wrote:
@Hammer

I'll put it another way. Arm won't die if Apple decide to switch to a different architecture in the future.

ARM can die at Softbank's hands.

From https://www.semianalysis.com/p/arm-changes-business-model-oem-partners

Quote:


According to Dylan Patel of SemiAnalysis, who examined court documents, Arm will reportedly change terms to use its IP where the use of other IP mixed with Arm IP is prohibited. If a chip maker plans to use Arm CPU IP, they must also use Arm's GPU/NPU/ISP/DSP IPs. This would result in devices that utilize every design the UK-based designer has to offer, and other IP makers will have to exclude their designs from the SoC. By doing this, Arm directly stands against deals like the Samsung-AMD deal, where AMD provides RDNA GPU IP and would force Samsung to use Arm's Mali GPU IP instead. This change should take effect in 2025 when every new license agreement has to comply with new rules.



The worst nightmare for ARM's TLA customers and guerilla marketing from ARM against companies like Qualcomm.

Note the momentum towards RISC-V and Apple's LLVM CPU insurance.

From Dylan Patel,
----------------
According to the updated Qualcomm counterclaim, after 2024, Arm is no longer going to license their CPUs to semiconductor companies such as Qualcomm under technology license agreements (TLAs). Instead, Arm will only license to the device-makers. Arm is allegedly telling OEMs that the only way to get Arm-based chips will be to accept Arm’s new licensing terms. Qualcomm claims that Arm is lying to Qualcomm’s OEM partners about Qualcomm’s licensing terms.

Furthermore, Qualcomm claims that Arm is telling the OEMs that semiconductor manufacturers will not be able to provide other elements of their Arm-based SOCs that Arm also offers as a licensed product. This includes GPUs, NPUs, and ISP. It seems that Arm is effectively bundling its other IP with the CPU IP in a take-it-or-leave-it model. That would mean Samsung’s licensing deal with AMD for GPU or Mediatek with Imagination GPU is no longer allowed after 2024. Furthermore none of these firms could use their in-house ISP or NPU despite it being far superior to Arm's.

If true, it seems Arm is playing very dirty with their threats to Qualcomm and OEMs. Mediatek, Samsung, and other Arm partners should be very scared. This is going to accelerate RISC-V roadmaps rapidly. It also reeks of anti-competitive behavior. Nvidia has a 20-year Arm license secured, so they will be fine. Apple obviously has great licensing terms due to their history with founding Arm. We hear Broadcom also has very favorable terms as well.

If you don’t believe us and think this all seems crazy, we get it. This is what we understand after reading the legal document. Just read this excerpt from Qualcomm’s counterclaim that was filed to the courts yesterday. We have attached the document at the end of the article.

--------------

https://www.youtube.com/watch?v=3D8TEJtQRhw
"I Am Altering the Deal, Pray I Don’t Alter It Any Further." - Vader

Amiga platform should maintain 68K CLR.

Softtbank's mismanaging ARM might end up like Japan Inc's SuperH.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

 Status: Offline
Profile     Report this post  
cdimauro 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 30-Oct-2022 6:54:09
#199 ]
Elite Member
Joined: 29-Oct-2012
Posts: 3621
From: Germany

@matthey

Quote:

matthey wrote:
kolla Quote:

Talking about something isn’t the same as having a plan. I mean, Gunnar has been talking about ASIC since forever. And you yourself talk about 68k ASIC pretty much constantly, but there’s hardly a plan.


There is too much uncertainty to develop a bigger plan. Uncertainty is the killer of business plans. It makes risk assessment too hard. The same is true for Amiga Corporation. Some people think Amiga Corporation should do more which they could but they are being conservative and maintaining financial discipline. A good example is when Michele Battilana was asked about an Amiga Maxi. They may have discussed plans with business partners but there is too much uncertainty to move forward despite opportunities for the Amiga.

This tells a lot. It's a big risk. So, business men tend to avoid big investments.
Quote:
In contrast, Hyperion has gambled and gambled on their gambles. Go to the casino, keep betting everything and it won't take long to lose everything.

Then it's their problem, right?
Quote:
cdimauro Quote:

Why not? Youngs are used to play games on any possible device.


With PPC hardware, there is no hardware close to being competitive in performance/price while younger people often have less money and are thrifty.

PowerPC was never competitive on that aspect.
Quote:
With emulation, there is a leaning curve and there may be a lack of ROMs while the gaming quality can be low due to lag/latency. THEA500 Mini did a nice job of presentation with minimal out of the box configuration so it is possible to do better. It should already be possible to distribute an emulator+ROMs with licensing for an auto running Amiga program in a 68k Amiga VM.

Michele's Amiga Forever succeeded at it: install, run it, chose the game from the frontend, and play.
Quote:
It's never going to be as attractive as native competitive hardware though.

Matt, I don't think that the people which are using an A500Mini would understand the difference. They just plug & play. That's it.
Quote:
cdimauro Quote:

For the 68020 isn't that much important, because at the time (Amiga 1200 & 4000) coders were a bit more aware of processors differences.

But for the 68000 it can be very important, unfortunately. It's strange that there's no cycle-exact 68000. AFAIR the core in WinUAE does it, so that it should be enough to copy the same implementation.


I believe it would be possible to make a cycle exact 68000 in FPGA although it would be much more work. It is surprising it hasn't been done by Atari fans but this likely an indication of how compatible the TG68 is without cycle exact timing.

It might be.
Quote:
I have doubts about how cycle exact emulation can be. It is difficult for software code to get down to the granularity of a single cycle.

It's more complicated but possible. You've to split the single internal operations like Lego blocks, organize the emulator as a raster line + Color clock base, and call each "Lego block" according to the specific slot conditions.

 Status: Offline
Profile     Report this post  
V8 
Re: Trevor Dickinson nominated for comment of the year!
Posted on 30-Oct-2022 7:36:39
#200 ]
Regular Member
Joined: 30-Mar-2022
Posts: 129
From: Unknown

@agami

Quote:
$4M is what I have extracted based on my interactions. Ben claims $10M.


Nonsense. The costs are virtually zero. Hyperion has never paid any of their developers
and that is why they, the developers, are no longer developing.

4M is possibly the amount of money that Ben has pocketed personally by providing "legal services" to hyperion.
Funny that, how almost all revenue is just transferred straight into Ben's pocket.

Just like how Trevor's money in AENv1 was also transferred straight into Ben's pocket.
Trevor's mistake was to not take Ben to court and instead ask for a "free OS4 licence for all my platforms going forward". To which Ben replied "Sure, and I will just never develop OS4 for Tibor. Checkmate."
This is why you never ever under any circumstances sign a document with Ben.

Last edited by V8 on 30-Oct-2022 at 07:41 AM.

 Status: Offline
Profile     Report this post  
Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 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