I'm looking for benchmarks comparing the various retro hardware Amigas out there i.e.
A500 Mini, V4SA, A1200NG, A600GS, Mister has anyone done that?

_________________

@Mobileconnect
Quote:

has anyone done that?

Most are only presenting Sysinfo screenshots, and this is one of the worst bemchmark programs out there.

Raspberry Pi 3 running Amibian:
[img align=left]https://hothardware.com/Image/Resize/?width=1170&height=1170&imageFile=/contentimages/Article/2639/content/big_amiga-sysinfo.jpg[/img]

Vampire4 SL:


The A1200NG:


A500Mini:

Last edited by OneTimer1 on 14-Nov-2025 at 11:09 PM.
Last edited by OneTimer1 on 14-Nov-2025 at 11:06 PM.
Last edited by OneTimer1 on 14-Nov-2025 at 11:04 PM.
Last edited by OneTimer1 on 14-Nov-2025 at 11:03 PM.
Last edited by OneTimer1 on 14-Nov-2025 at 10:58 PM.

@Mobileconnect

Quote:

I'm looking for benchmarks comparing the various retro hardware Amigas out there i.e. A500 Mini, V4SA, A1200NG, A600GS, Mister has anyone done that?

None of those are actually retro hardware.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

And SysInfo is pure crap as benchmark.

Why people don't use:
- Doom for general purpose code;
- Quake for a mixture of general purpose and floating point code;
- Lightwave for floating point intensive code
or other equivalent applications?

@cdimauro

Quote:

And SysInfo is pure crap as benchmark. Why people don't use: - Doom for general purpose code; - Quake for a mixture of general purpose and floating point code; - Lightwave for floating point intensive code or other equivalent applications?

Sysinfo is one single program generating nice graphics, without the need for installation of different real world applications

And it is crap because it is mainly using short loops, that are running best in systems with caches or if you are using a JIT like mechanism.

It ignores differences in Chip-RAM / Fast-RAM / ROM speed, FPU or HD access.
A Vampire4 SL may have a better FPS rate on GLOOM than the A1200NG but Sysinfo will shine with JIT translated the short benchmark routines.

Last edited by OneTimer1 on 15-Nov-2025 at 08:56 AM.

@cdimauro

I have noticed everyone who benchmark stuff, is only interested in fake scores

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

Thanks everyone who reply, except kolla, who's opinion I do not care about

Thanks especially for the screenshots from sysinfo albeit i know sysinfo is not the be all of accurate benchmarks. But it can suffice when its the same test applied to all.

As for the debate about what benchmarks are valid or not, I also do not care about that I am only interested in an empyrical comparison of those various re-implementations of Amiga's performance. if you want to debate which benchmark is most accurate make your own thread or give me a definitive answer with proof.

_________________

@Mobileconnect

People used to grumble about SysInfo not being accurate back in the day when we were all using actual Commodore Amigas, but back then it gave a good ruff idea of the real performance.

As kolla pointed out "None of those are actually retro hardware". The fact that your asking about modern FPGA/Emulation based Amigas rather than retro (based on a real physical Motorola 680xx (ASIC) based systems) hardware goes to the heart of why SysInfo is an extremely poor choice for comparisons.

OneTimer1's post #5 explains why its now so bad ... I'm no expert but as I understand it you can ended up with SysInfo giving you the impression the system that's the fastest in some real world software tests is slower than another system that is in reality slower!

_________________

@Tpod

Quote:

Tpod wrote: People used to grumble about SysInfo not being accurate back in the day when we were all using actual Commodore Amigas, but back then it gave a good ruff idea of the real performance.

Ever tried AIBB (read the full story here https://www.edsa.uk/blog/the-emu-and-the-storm and stay away from AOS3.2)?
It is a bit like SysInfo but gives more precise information:

@OneTimer1

AIBB is relatively new to me in Amiga terms; first used it about 15 years ago whereas I first used SysInfo way back around 1992.

AIBB gives a lot of different test options (helps get a better idea of the real system speed) but SysInfo gives other nicely presented info which can also be useful, I like them both. I guess even AIBB wouldn't be ideal when it comes to the modern gear Mobileconnect is interested in though.

Last edited by Tpod on 16-Nov-2025 at 12:54 AM.

_________________

@Tpod

Quote:

Tpod wrote: AIBB is relatively new to me in Amiga terms; first used it about 15 years ago whereas I first used SysInfo way back around 1992.

It was common in the mid 1995 when I had a lot to do with Amigas, what really counted where the tools that help to discover differences in Accelerators.

There where some that had good clock speeds but miserable ChipRAM access speeds, some failed when using big amounts of FastRAM or had no KickROM or the RTG systems sucked.

I was supriesed to see a difference between AOS3.1 and AOS3.2 that was slower with screen routines.

Quote:

Tpod wrote: AIBB gives a lot of different test options (helps get a better idea of the real system speed) but SysInfo gives other nicely presented info which

Nicely presented info which can be misleading.

@Mobileconnect

There's a LightWave test here.

Classic machine results range from hours to around 30 minutes for an overclocked 060.
Pistorm32 / RPi4 is in single figure minutes.

My Raspberry Pi 5 (2.8GHz) running Amiberry did it in 47 seconds.

Last edited by minator on 16-Nov-2025 at 02:03 AM.

_________________
Whyzzat?

best real world application for benchmarks

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

@OneTimer1

Quote:

OneTimer1 wrote: @cdimauro Quote: And SysInfo is pure crap as benchmark. Why people don't use: - Doom for general purpose code; - Quake for a mixture of general purpose and floating point code; - Lightwave for floating point intensive code or other equivalent applications? Sysinfo is one single program generating nice graphics, without the need for installation of different real world applications And it is crap because it is mainly using short loops, that are running best in systems with caches or if you are using a JIT like mechanism. It ignores differences in Chip-RAM / Fast-RAM / ROM speed, FPU or HD access. A Vampire4 SL may have a better FPS rate on GLOOM than the A1200NG but Sysinfo will shine with JIT translated the short benchmark routines.

That's why it shouldn't be used: it's misleading (at least).


@NutsAboutAmiga

Quote:

NutsAboutAmiga wrote: @cdimauro I have noticed everyone who benchmark stuff, is only interested in fake scores

Self-satisfaction of its own toy: a psychological attitude.

With a bit of... my schwartz is bigger than your:

@Thread


Most people are more in games.

Older Amiga games don't benefit much faster CPU/GFX speed.
Most Games that where especially made for later Amigas do benefit from CPUs up to 030/50, good ChipRAM performace andd some are supporting RTG.

Only productivity Software and Ports will benefit from faster systems.

@NutsAboutAmiga

> I have noticed everyone who benchmark stuff, is only interested in fake scores

People are paying for performance, therefore bench marking seems important.

It might be more interesting if 'new retro' hardware is coming with I/O software and drivers.

Ask yourself:

Does it work out of the box or do I need to spend days of configuration?
Do I want a system that is open for installation of Amiga software?
Do I need a system where I could play games from (W-)LAN / SD-Card / USB-Stick?
Do I need an Amiga case with keyboard

Last edited by OneTimer1 on 16-Nov-2025 at 11:05 AM.
Last edited by OneTimer1 on 16-Nov-2025 at 10:55 AM.
Last edited by OneTimer1 on 16-Nov-2025 at 08:20 AM.

@OneTimer1

Quote:

OneTimer1 wrote: Quote: Tpod wrote: AIBB gives a lot of different test options (helps get a better idea of the real system speed) but SysInfo gives other nicely presented info which Nicely presented info which can be misleading.

The SYSTEM SOFTWARE INSTALLED & DRIVES Information is the nicely presented & useful info I was referring to; I'm not aware of any of that being misleading.
ShowConfig in OS3.2.3 now gives a lot of the same info which is also nicely presented . Do you know how accurate the DRIVES Speed Test results are?

_________________

@kolla

Quote:

kolla wrote: @Mobileconnect Quote: I'm looking for benchmarks comparing the various retro hardware Amigas out there i.e. A500 Mini, V4SA, A1200NG, A600GS, Mister has anyone done that? None of those are actually retro hardware.

Actually they are.

Retrocomputing
Quote:

Retrocomputing is the current use of older computer hardware and software... Some retrocomputing enthusiasts also consider the "homebrewing" (designing and building of retro- and retro-styled computers or kits), to be an important aspect of the hobby, giving new enthusiasts an opportunity to experience more fully what the early years of hobby computing were like. There are several different approaches to this end. Some are exact replicas of older systems, and some are newer designs based on the principles of retrocomputing, while others combine the two, with old and new features in the same package.

Real hardware from the period when the Amiga was being produced by Commodore is more accurately described as 'vintage', whereas 'retro' includes modern recreations and upgrades. What makes it 'retro' is operating similarly to the original hardware and software.

The Amiga stretches this further than most other vintage computers because it was designed to be extended with virtually no limit. In the 80's and 90's we were shoving the fastest 68k we could get into them, and then PPC in the late 90's. Also higher resolution graphics and 16-bit sound chips, integrated into the OS.

All of this is retrocomputing, even when the accelerator is a PiStorm or Vampire. Mister is an FPGA reproduction of an entire Amiga which also qualifies, as is V4SA.

Software retrocomputing (aka emulation) is also a thing, but this isn't hardware so doesn't apply to this thread. You could argue that theA500 Mini is not retro hardware because the entire machine is emulated in software, but it does have custom hardware that acts (sort of) like an A500.

TheA500 Mini, V4SA, A1200NG and A600GS are all designed to be standalone replacements for vintage Amigas, able to run 'vintage' software as well as modern 'retro' titles. The relative performance of these machines could be of interest to people trying to decide which system(s) they might want, whether for practical reasons or just for bragging rights (a valid retro activity that continues a tradition which started back in the 80's).

@bhabbott

Quote:

bhabbott wrote: The Amiga stretches this further than most other vintage computers because it was designed to be extended with virtually no limit. In the 80's and 90's we were shoving the fastest 68k we could get into them, and then PPC in the late 90's. Also higher resolution graphics and 16-bit sound chips, integrated into the OS.

What about PCs, then?

emulator is emulator and nothing more than emulator
want to use emulator use uae on pc
it is fastest and cheapest solution
uae was fast and cheap in year 2000 and still is best choice
no reasons to waste time on benchmarks on arm

Last edited by ppcamiga1 on 23-Nov-2025 at 07:38 AM.
Last edited by ppcamiga1 on 23-Nov-2025 at 07:38 AM.

@ppcamiga1

Quote:

ppcamiga1 wrote: emulator is emulator and nothing more than emulator want to use emulator use uae on pc it is fastest and cheapest solution uae was fast and cheap in year 2000 and still is best choice no reasons to waste time on benchmarks on arm

Then be coherent and disable Petunia and RunInUAE from your OS4 setups...