The “Cycle Accurate” Conundrum and Is There a List of Cores That Are?

[Syntax Error]

Hi,

(the eventual question is : "is there a list of cores, both console and computers, that are deemed "cycle-accurate" and what exactly is meant by that denomination in this case)
i cant say im the greatest expert on FPGA and its programming languages but i actually got one b/c it had an amiga and atari st/ste in it, promising in the future the possibility of a jaguar and a falcon.
For about a week now (its quite a lot to take in i dare say). At first glance everything looked a lot more straightforward than i thought it would be, by now i understand getting final fantasy 1 to run without 2cm top + 2cm bottom tearing on my HD screen doesnt come by itself (but it works).
Now, the point, i slowly start reading the forum here and there and the first things that stand out here is both amiga <i>and</i> atari cores claim to be not cycle-accurate but to just run most demos.
This somewhat defies the purpose of forking out the money and not just using an emulator tho i dont regret getting it and its the cheapest solution i know off that holds a world of emulated emulators (if its not a perfect copy of the machine) in it running at 5v/2a (mine so far) 10 watts consumption.
Since i would be mostly interested in the computers and actually programming a little this sounds like it might be a problem.
For instance the latest install of VICE for c64 claims to be but i still saw stuff on csdb where people claimed some .d64 "probably wasnt tested on real hw" and therefor didnt work on it, but ran perfectly on the emulator. I have Hatari on linux and im sure i can get the popular winUAE to work but i also dont know how exact those would be. The main thing is i dont have space (and definitely money to get a real a1200 / Falcon / St(e) in the house)

So ( ... its a mental condition where i keep elaborating i think) : Is there a list of cores, both console and computer (dont really care about the arcade cores as long as the games there work) that are considered "cycle-accurate" and what exactly does the denomination in this case stand for ? For instance in case of ST i read somewhere you cant use ".stx" format b/c the core is not cycle accurate, that means its a floppy disk that wont work but does that mean the computer itself isnt perfect or does that mean the implementation of the floppy drive isnt perfect or does that mean the transmission (of electrons ?) between the two systems within the real box isn represented perfectly by , ... lets call it "the ghost in the machine"

Its probably a tall order but i have time, i have a box that fits about one soap bar that has a multiverse inside and eats only 10 watts, generates no heat on top of a 14 cm case fan (extra) and makes no noise at all, no moving parts so it should last a while.
Any info or explanation on the matter is welcome, if not only a list of cores that are considered to be, more elaboration on the subject. Im trying to understand.

If anyone feels like making time to respond and / or explain thanks in advance,

J

[vteckickedinyo]

I believe there are only two cores that could be currently at the time of this writing considered accurate, NeoGeo & the new MegaDrive/Genisis core. (NUKED-MD) And yet, both cores are still being improved so obviously it isn't exactly functioning like the old hardware... (Yet, possibly maybe.)
I felt a bit the same way with MiSTer after run ahead mode came out for RetroArch but at the end of the day, a well equipped MiSTer still functions with less latency on average and there is less tinkering with latency on a per game level. I personally enjoy the low power usage of the MiSTer for what it does.

news.ycombinator.com, user: Pubby "...Cycle-accurate means the emulator is emulating all of the little steps that make up an individual instruction"
I believe the MiSTer excels at running instructions in parallel meeting timing cycles with lower frequency measured in Mhz, in which most software emulators (I think?) run each individual instruction one at at time but quickly in the Ghz range. I assume there are software out there that is counting on specific instructions to be done simultaneously.

Pulled from news.ycombinator.com, user: bcrl "...For example, the documentation for Amiga's custom chips specifies which cycles various DMA activity takes place. Replicating that in hardware is quite easy (just a few comparators wired up to a counter), but it's hugely expensive in software which leads to all kinds of nasty hacks to replicate behaviour games rely on."

So MiSTer will likely never be nanosecond accurate to the actual hardware, but it's likely the closest you're going to get other than to going out and getting the original hardware.

I'm sure I have said something wrong and someone else will pick up where I'm trailing off because I don't have much experience with the ST nor it's cores on MiSTer.

[Newsdee]

The term is thrown around but doesnt mean that much when talking of an overall system. Maybe the disk drive is not cycle accurate, but the CPU may be. That means some niche uses may not work but all games and demos work perfeclty anyway.

To me one thing that that matters is how close we are of accuracy for the power consumed. FPGAs beat hands down CPU emulators for most systems. And thats normal given their parallel nature (versus a CPU that runs everything sequentially). Runahead might reduce latency, but to me its just a brute force solution instead of using more approrpriate tech.

[Armakuni]

ZX Spectrum is one go add to list

NES/SNES ?

A lot of cores you wouldn't be able to tell if they are not though unless you know a system inside out and know what to look for

Some cores are held back by open source modules too like TG68k 68020 CPU

[Duffygag]

Do you know what does it mean by the way? Maybe you are asking a question without knowing the implications, why does or doesn't matter...

[rhester72]

No disrespect intended, but there is no "conundrum" and the definition of cycle-accurate has never changed, be it in software or FPGA emulation...it means the time of execution for each instruction takes the same number of cycles as on original hardware (and adjusted appropriately for various state conditions). It is an indication of accuracy only so far as having shorter or longer cycle counts negatively affect things like demos but really has very little to do with the perceived accuracy of emulation and certainly nothing whatsoever to do with FPGA.

All that being said, OP, it sounds like you bought into some sort of Tesla-like hype train with the MiSTer that claims "doing it in hardware" is somehow inherently better...which it is not.

Sadly, with byuu/Near's passing, his original definitive article on the subject has been lost, but a lot of commentary around it can be found here:

https://www.reddit.com/r/emulation/comm ... reator_of/

which gets to the root of it. More or less, the following is true:

• There's pretty much nothing you can do in FPGA that you can't do in software, except efficient mass-scale parallelization
• Most of the cores you find in FPGA are adapted from software implementations (MAME, et al) and largely inherit the same limitations and implementation shortcomings as their software brethren
• Any of the things that lead to higher emulator accuracy - decapping of original chips for more precise understanding of original hardware construction, measurement and implementation of accurate cycle counts for instruction execution, better timing integration with external video/audio/IO chips, more accurate implementations of filters, etc. are equally valid and possible in software or FPGA
• FPGA is inherently limited by two factors - available logic elements and switching speed - that can generally be overcome by simply throwing more compute power at it in a software implementation that becomes extremely prohibitively expensive in FPGA (the economy of scale just isn't there)

The short version: If you bought MiSTer because "hardware is better than software" where it comes to emulation, you were misled.

That doesn't make MiSTer any less amazing. Fully assembled, it's a box the size of three decks of playing cards, and yet brings MASSIVE levels of quality emulation of arcades, consoles, and computers together in a single package with a consistent user interface that is generally equally fantastic on HDMI or old-school CRTs for proper 'feel' (something much more challenging on most modern PCs). It can in some cases directly interface with other OG hardware (SNAC, for example) for experiences difficult to duplicate (at least in terms of latency) on other solutions. It truly is a tiny miracle in a box absolutely worthy of all the love, passion and interest it has attracted over the years...but it is not perfect, and never will be.

Like other emulation platforms, MiSTer strives to achieve a midpoint of convenience versus (perceived) accuracy to allow for enjoyment of as much retro content as possible. It's not the only solution, it's not the "best" solution (because there isn't one)...it's another (sadly rather more expensive now) solution, and is more optimal for some than others depending on exactly what you need from it.

If your goal is 100% accuracy at the expense of all other considerations, I'm afraid nothing but original hardware will do. For most, myself included, the convenience offered by a "98% accurate" (I'm just throwing that number out there, it means nothing) solution in a portable package that works with CRTs and offers super-low latency, it's a worthy tradeoff versus owning several hundred arcade cabinets, consoles with cartridges and discs, and computers with masses of floppy disks stored in several inconveniently-large storage lockers.

As always, yours (and everyone's) mileage may vary, but I wanted to do my best to ensure you have a reasonable understanding of what MiSTer is...and isn't.

[dshadoff]

As one digs deeper into compatibility, one finds that "cycle accurate" is measuring really only one set of variables related to compatibility.
There are edge cases related to register initialization values, and interactions between any two components inside the machine - which means that more components means exponentially larger numbers of edge cases.

In short, if you are looking for an "identical match to hardware", you won't find one on FPGA, nor will you on software. Heck, on many systems, you won't even find uniformity between hardware revisions. On most systems, the sum total of what is known about each component is not complete enough to even create a provably-complete set of tests for all of this (although subsets do exist for some machines).

Perhaps the goal that people should be seeking is how many known compatibility issues there are with software released on the system, or whether the core is "more compatible" than any other software or hardware implementation available. For example, the PC Engine (TurboGrafx) core is the best implementation of the system, with a vanishingly-small number of issues, of which almost none affect any commercial game.

[Armakuni]

No disrespect intended, but there is no "conundrum" and the definition of cycle-accurate has never changed, be it in software or FPGA emulation...it means the time of execution for each instruction takes the same number of cycles as on original hardware (and adjusted appropriately for various state conditions). It is an indication of accuracy only so far as having shorter or longer cycle counts negatively affect things like demos but really has very little to do with the perceived accuracy of emulation and certainly nothing whatsoever to do with FPGA.

All that being said, OP, it sounds like you bought into some sort of Tesla-like hype train with the MiSTer that claims "doing it in hardware" is somehow inherently better...which it is not.

Sadly, with byuu/Near's passing, his original definitive article on the subject has been lost, but a lot of commentary around it can be found here:

https://www.reddit.com/r/emulation/comm ... reator_of/

which gets to the root of it. More or less, the following is true:

• There's pretty much nothing you can do in FPGA that you can't do in software, except efficient mass-scale parallelization
• Most of the cores you find in FPGA are adapted from software implementations (MAME, et al) and largely inherit the same limitations and implementation shortcomings as their software brethren
• Any of the things that lead to higher emulator accuracy - decapping of original chips for more precise understanding of original hardware construction, measurement and implementation of accurate cycle counts for instruction execution, better timing integration with external video/audio/IO chips, more accurate implementations of filters, etc. are equally valid and possible in software or FPGA
• FPGA is inherently limited by two factors - available logic elements and switching speed - that can generally be overcome by simply throwing more compute power at it in a software implementation that becomes extremely prohibitively expensive in FPGA (the economy of scale just isn't there)

The short version: If you bought MiSTer because "hardware is better than software" where it comes to emulation, you were misled.

That doesn't make MiSTer any less amazing. Fully assembled, it's a box the size of three decks of playing cards, and yet brings MASSIVE levels of quality emulation of arcades, consoles, and computers together in a single package with a consistent user interface that is generally equally fantastic on HDMI or old-school CRTs for proper 'feel' (something much more challenging on most modern PCs). It can in some cases directly interface with other OG hardware (SNAC, for example) for experiences difficult to duplicate (at least in terms of latency) on other solutions. It truly is a tiny miracle in a box absolutely worthy of all the love, passion and interest it has attracted over the years...but it is not perfect, and never will be.

Like other emulation platforms, MiSTer strives to achieve a midpoint of convenience versus (perceived) accuracy to allow for enjoyment of as much retro content as possible. It's not the only solution, it's not the "best" solution (because there isn't one)...it's another (sadly rather more expensive now) solution, and is more optimal for some than others depending on exactly what you need from it.

If your goal is 100% accuracy at the expense of all other considerations, I'm afraid nothing but original hardware will do. For most, myself included, the convenience offered by a "98% accurate" (I'm just throwing that number out there, it means nothing) solution in a portable package that works with CRTs and offers super-low latency, it's a worthy tradeoff versus owning several hundred arcade cabinets, consoles with cartridges and discs, and computers with masses of floppy disks stored in several inconveniently-large storage lockers.

As always, yours (and everyone's) mileage may vary, but I wanted to do my best to ensure you have a reasonable understanding of what MiSTer is...and isn't.

Even with a more soft approach to core development the results can be better on FPGA and we have cores that are timing accurate without being cycle accuarte, That is not saying FPGA is a magic bullet of course

Original hardware wont be a solution forever on some platforms and we have seeing more and more drop in FPGA and ARM based replacements for original chips especially on some 1980s systems like the C64, so will come to point where a complete FPGA core is more efficient solution

On the majority of cores in a blind side by side test you would be hard pressed to tell the difference and that is what most people want a good enough solution

Software is always going to be the solution where FPGA ends especially when GPUs enter the mix as then there is far more variance in the what is done per cycle due to frame to frame times

[Xbytez]

Sadly, with byuu/Near's passing, his original definitive article on the subject has been lost, but a lot of commentary around it can be found here:

https://www.reddit.com/r/emulation/comm ... reator_of/

For those that wish to read byuu/Near's article, I've posted this archived version of the original here for ease of reading as the reddit links are now dead.

FPGAs Aren't Magic

Note: this article was originally written on February 6th, 2018. I tried to present an unbiased look that expresses the pros and cons of each approach, but it seems that the underlying message was lost.

The original intent of this article was to express that whether coding in Verilog for FPGAs, or C++ for PCs, these languages are both Turing-complete. That is to say, there is absolutely nothing that one can do that the other cannot. It's only a matter of computational resources and energy cost.

The reason for writing this article originally was that I felt software emulators were given an undue bad rap for the purpose of marketing a product. I can state objectively that software emulation today is more accurate than the best FPGA cores. In fact with new projects such as MiSTer, a more collaborative open-source approach to FPGA emulator development has emerged, and we are now combining our resources and sharing our work.
I am a proponent of both FPGAs and software emulators. All I ask is that each approach is given the respect they each deserve.

Original Article

c03b2eefc5d5d879dce5e7b137e2a2bf86a0b667.jpg (88.08 KiB) Viewed 14341 times

Recently, there's been a lot of news hype around the Analogue Super Nt. And unfortunately, sites like Polygon are putting out glorified press releases instead of reporting factual information. So I'd like to dispel some dangerous misrepresentations now before they spread further.

Code: Select all

• FPGA devices are emulators too
• nothing inherently makes FPGAs more accurate than software emulation
• latency is caused by host operating systems, not by choice of programming language
• the Analogue CEO is selling you a product; so expect hyped marketing claims

As immediate evidence that the Super Nt is not perfect, the initial hardware release ships with a bug that causes Rendering Ranger to be unplayable. I am sure it will be fixed soon, but it demonstrates right away that the Super Nt is not a perfect clone of the SNES.

Foreword

I have nothing against the Analogue Super Nt. I think it's a great product. In fact, I'll start by conveying by what I perceive as its benefits.
Over software emulators:

Code: Select all

• reduced latency as a result of being a single-purpose device with no OS overhead
• lower power consumption at the same degree of accuracy as a software equivalent
• lower cost of ownership if one does not already own a capable PC
• instant-on capability

Over real hardware:

Code: Select all

• HDMI output bypasses the need for expensive XRGBs or analog-RGB CRTs

But it also has its limitations:

Code: Select all

• it costs $190; about five times what a real SNES goes for
• it runs at 60hz instead of the proper 60.09hz of a real SNES
    • it can run at 60.09hz, but only with either frame-dropping or tearing
• no support for video filters, save states, real-time rewind, netplay, etc
• playing homebrew, hacks, translations, MSU1 games, etc requires an expensive flash cartridge
• cannot be used as a debugger for homebrew, ROM hacks, speedruns, etc
• if one has a PC: this takes up additional space, a power outlet, and an HDMI port

And of course, higan has its own limitations, in addition to the four things above which the Super Nt does better, it also:

Code: Select all

• is incredibly slow
• has a complex user interface
• requires copyrighted firmware to run games containing DSP coprocessors
• is missing some software features of other emulators (netplay and rewind)

The way I see it: there's no such thing as perfection. Everything has its pros and cons. And everyone is going to have different priorities and values. It's a good thing that both the Super Nt and higan exist. We can coexist cooperatively instead of competing with one another.

Emulation

60b76d91e19b374b466562d13dea9cf73d656bc0.png (22 KiB) Viewed 14341 times

All emulation works the same way: we study the original hardware, read documentation and datasheets, develop test code to run on hardware to verify our claims, and then implement the results in source code.

The end product attempts to provide a faithful recreation of the original hardware. The accuracy of this recreation is bounded by raw processing power: that is to say, more accurate emulation requires more resources.

There are those that will take shortcuts, and Chrisopher Taber, the CEO of Analogue, brought those emulators up in the Polygon article.
But absent from the article was any mention of software emulators that are serious about emulation: projects like higan (SNES), BlastEm (Genesis), gambatte (Game Boy Color), mGBA (Game Boy Advance), etc.

Instead, you have the following quotes:

"No matter how hard the (software emulator) developers try, the games will never sound exactly the same as the browning lump of plastic in your closet. Slight imperfections mar even the best emulators: the audio will bleep out for a second, or a sprite will flicker when it should fade."
"Mention the word "emulator" to him, and he'll immediately rev up his proselytizing process."

These quotes are deeply unfair. higan currently has zero known emulator bugs, despite us having tested the entire SNES library with it. No audio drop-outs, no sprite flickering instead of fading, nothing.

There are no examples provided in the article pointing to any issues with higan, because for more than a decade, I've fixed any and all reported SNES emulation bugs within days of them being reported to me.

Of course, it's not perfect. And neither is the Super Nt. You see, the Super Nt is not a real SNES. It is not a perfect transistor-level replication of the original hardware. It is prone to the same bugs, and goes through the same bug-fixing and improvement process as higan does. The Super Nt will never be the same as the browning lump of plastic in your closet, either. Yet this isn't mentioned in the article.

FPGAs

FPGAs, or field-programmable gate arrays, are components that are programmed through code written in languages such as Verilog or VHDL. This code tells the component how to operate, in the same way that C++ instructs a general purpose CPU how to operate.

How these two differ behind the scenes (logic gates versus microcode) is irrelevant to the end result: replicating the original hardware experience. Setting efficiency aside, there is absolutely nothing an FPGA can do that cannot be done in software. And indeed, there is nothing the Super Nt does that higan does not.

FPGAs may be inherently more parallel, which is beneficial for resource usage, but modern CPUs are really, really good about context switching. Software emulators can and do verifiably simulate down to the tiniest possible clock cycle, the parallel nature of real hardware simply by using more resources.

If you believe that HDLs such as Verilog are magic, consider that automated translators exist to transform C code into it. They just don't tend to operate as well in larger-scale applications. Yet.

In both cases, emulator developers are transforming their knowledge of how the original SNES works into source code that attempts to mimic their operation. Given equal knowledge of the original hardware and absolutely masterful potential, an FPGA implementation and a PC software implementation will be equally accurate.

In the case of higan, I've spent thirteen years working full-time on increasing my understanding of the SNES hardware, and refining my emulation of it. I've worked with dozens of people over the years to help me in this goal. The Super Nt has been in development for roughly one year. Furthermore, I consider its lead developer, Kevin Horton, a friend. We work together and share information. I'm in the credits on the Super Nt device for a reason: emulation is built on the shoulders of giants. The Super Nt is possible in part by higan, which was possible in part by Snes9X, etc. And indeed, the relationship is mutually beneficial: findings from Kevin have already made their way into higan.

An emulator is not measured by any one bug or lack thereof: it's the total sum of years of testing, validation, and tens of thousands of unique behaviors all operating together correctly in harmony. These things take time.

I don't want to write an article critical of the Super Nt itself, nor of its lead engineer. But I am very distressed at how Chris' words in this story paint all software emulators as inferior to his product, ostensibly for marketing reasons. Even if that isn't his intention, that's how this article reads.

Latency

The one area where the Super Nt will absolutely beat traditional emulators is in latency: how long it takes between when you press a button on the gamepad until you see and hear the result of said action on the screen.

The reason for this is once again not magic: the Super Nt runs without an operating system in the way. Yet when you run an emulator on your desktop, it has to share resources with a thousand other processes that also want access to your video card, your sound card, your input devices, etc. This time sharing results in added latency. A software emulator can reasonably expect to get within 30-50ms of the latency of a pure hardware approach.

But again, it's not magic: there is nothing preventing an emulator written in C from running on bare metal, without an operating system in the way. It isn't done only because the demand isn't there to produce a robust real-time kernel environment that gives software emulators direct ownership access to all hardware resources. And thus, emulator developers cannot bypass the need to share these resources. Yet.

But indeed, if latency is your primary concern, I concede that FPGA devices are currently the way to go. Software emulation latency is almost imperceptible these days with the advent of adaptive sync, WASAPI/ASIO, 1000hz USB polling, etc, but it cannot be eliminated so long as users wish to run emulators right alongside their web browsers and photo editors.

The best software emulators claim to reach within 8ms of real hardware latency, and with all the above in higan I myself can perceive no difference to my real hardware, however I do not have the equipment to verify such claims, so I will leave that to another discussion.

Preservation

Chris is eager to tell you a story about his deep love for retro gaming. And I believe he is sincere.

But he is not the first to take SNES emulation accuracy seriously: I've been doing that since 2004.

6ea403525787cb7f94ecd436e86961f007f98dc8.jpg (45.61 KiB) Viewed 14341 times

But he is not in it to preserve the hardware for future generations: he's in it for the profit. His product is necessarily closed source. When the last Analogue Super Nt dies of old age, it will be a distant memory. It is a black box, just like the original SNES consoles.

I'm not going to begrudge someone for wanting to earn a living. But you need to understand: Polygon's article is nothing more than a product advertisement, devoid of any and all actual journalism and fact-checking. I am going to take issue with my life's work being misrepresented, as it has been here.

Game Preservation

These clone consoles that take physical cartridges bypass the need to emulate the hardware that exists inside of game cartridges. Game cartridges will one day fail just as game consoles will. And when they do, the Super Nt will not help you play SuperFX games like Yoshi's Island, or SA-1 games like Super Mario RPG.

Software emulation preserves the complete environments in digital form, not just the base systems. Those are meaningless without any content for them.

Technical Documentation

A very important point in fairness to mention is that Kevin Horton has promised to release detailed timing documentation on his findings while developing the Super Nt product.

THAT is preservation. And that is extraordinarily commendable, and we should all be deeply appreciative of Chris, Kevin, et al when this happens.
Without the Super Nt, we wouldn't have this documentation in the future.

Closing

Yes, I wrote this piece because the Super Nt press intentionally neglects my own work. And on that note, here's the bottom line: The Super Nt is not more accurate than higan is. And if you claim otherwise ... then I say to you, prove it.

This is also not to say that higan is more accurate than the Super Nt is! Only time will tell. And given the personalities of Kevin and I, it's extremely likely that the moment any issue is observed in either of our products, we'll be fixing it as soon as possible. And hopefully, helping each other with that.
But this truly isn't just about me: it's about the inevitability of seeing this claim that FPGAs are inherently superior in accuracy to software emulators repeated for more systems in the future if we don't set the record straight now.

The Super Nt is a fine product, produced by a fine engineer: but please stop with the grandstanding accuracy boasts. They're patently false, devoid of any evidence of said claims, and they're disparaging to the people doing the real work of preserving our gaming history.

[rhester72]

@Xbytez Thank you very much indeed for that

[thorr]

This is all well said. For me personally, the MiSTer is superior for these reasons, many of which were already mentioned above:

• SNAC: I can use real controllers with it including obscure ones like light guns with no lag.
• Latency: There is little to no overhead depending on the situation. In the case of NES, I can play Duck Hunt because there is no input, display, or other latency. Until there is an OS-less software emulator solution, FPGA is where it is at for this.
• Configuration: In recent days, I have spent many hours and different video cards trying to get a PC working the way I want it to with my arcade cabinet and due to various hardware and software compatibility and configuration issues, I so far have not succeeded. I keep having to research my issues and test changing settings and config files and work my way through all the problems. I finally have an older Radeon video card that is working with new Ryzen, but when I try to launch a game it freezes the screen on the CRT. It works on the LCD screen. Crap like this is what I am talking about. Configuring a MiSTer is immeasurably easier than what I have been going through, and I administer Linux daily at my job.
• Easily works with CRT's in the native resolution and video timing. This is possible in software, but it is a PAIN IN THE BUTT in my experience to learn about and configure and get working.
• Size: MiSTer is smaller than almost any powerful PC.

MiSTer vs. real original hardware - I prefer MiSTer because it has many advantages over original hardware:

• I can use component video. I modded my Atari 2600 with Svideo and it looks terrible in comparison to the MiSTer. Both play like the real thing with no lag. Why would I want crappy video quality?
• Space. The MiSTer is tiny compared to pretty much all real original hardware. I don't need cartridges either which also take up space.
• Convenience. I have all my consoles in one place. I don't have to switch out cords or change inputs, etc. other than swapping out the controllers.
• Additional features. Save states, cpu overclocking, etc. are possible if you want them, but are not required. You don't have that option on real original hardware.

Cycle Accuracy:

• I fully accept that MiSTer and software emulation is possibly not 100% accurate. I enjoy helping to find the bugs so they can be fixed.

[dmckean]

• There's pretty much nothing you can do in FPGA that you can't do in software, except efficient mass-scale parallelization

The importance of this can't be understated.

[dshadoff]

While it is true that everything which can be done on FPGA can also be done in software with a sufficiently-fast computer (*), I will remind people of a point I’ve made in the past:

In the early days of emulation, all system operation was inferred from either documentaion (i.e. 6502 datasheets), or software-based empirical tests. There are limits to the maximum precision of these tests, although they can create a “quite good” emulation model.

Hardware test equipment - needed to determine behaviour in circuit to a higher precision - was expensive, and out of the reach of most emulator developers. It is often easier to use this additional information in a hardware model than a software one, but the availability of the information can benefit both.

Likewise, decapping can provide an even finer-level of detail to the model of the original equipment, and answer many questions which could not be known otherwise. Again, the information gleaned from this can benefit both hardware and software emulation, provided that authors makemuse of it.

But this is where I’ve seen some divergence in recent years - software emulators are tending to add frills and features, and in many cases ignoring the newly-found compatibility information… in some cases, because the code has not been written in a way that it can be implemented easily. Again, drawing from my own personal experiences on PC Engine, some fixes for edge cases can easily be implemented in Mednafen, while others simply cannot. And I created my own personal fork of Mednafen, since the original author has not acted for several years on many of my compatibility information disclosures.

I try to keep all my compatibility discoveries updated in both the MiSTer core and my Mednafen fork, but many things simply can’t be updated in Mednafen without a rewrite of a substantial portion of the code (which I don’t have time for). For this reason, MiSTer is still much closer to the original hardware than the best software emulator (at this time).

(*) - provided that one exists, which has, historically, not always been the case.

[Stinky]

• There's pretty much nothing you can do in FPGA that you can't do in software, except efficient mass-scale parallelization
• Most of the cores you find in FPGA are adapted from software implementations (MAME, et al) and largely inherit the same limitations and implementation shortcomings as their software brethren
• Any of the things that lead to higher emulator accuracy - decapping of original chips for more precise understanding of original hardware construction, measurement and implementation of accurate cycle counts for instruction execution, better timing integration with external video/audio/IO chips, more accurate implementations of filters, etc. are equally valid and possible in software or FPGA
• FPGA is inherently limited by two factors - available logic elements and switching speed - that can generally be overcome by simply throwing more compute power at it in a software implementation that becomes extremely prohibitively expensive in FPGA (the economy of scale just isn't there)

For most software emulation means running on a general purpose computer system, with hardware shared between processes.This means abstracted I/O: inputs are read by the OS and passed on, display is rendered to a framebuffer, component synchronization has to be deterministic, hence the existence of compensating workarounds like runahead. With a platform like MiSTer the running core can interface directly to input devices through SNAC and directly drive the raster in a CRT, all things being equal the nature of an FPGA system can be more deterministically accurate than a PC emulator.

Not to say that PC emulation is not accurate or that software emulation could not be directly interfaced, and of course many people use USB inputs and HDMI output on MiSTer. But generally when we're talking software v FPGA this is what is being discussed. Also the opinion on economy is not entirely valid, the existence of MiSTer shows that advancing FPGA technology is becoming increasingly affordable over time.

[Newsdee]

What's missing from the "debate" (which is really marketing hype) is to clearly define what is being measured, and what do we mean by "better".

FPGAs are much better at having high throughput of signals (data transmission/conversion). That makes it great for replicating low level electronics, interfacing with CRTs, original peripherals. And for hitting cycle accuracy at the CPU level.

CPUs are better at doing things the real hardware could not do. Save states, rewind, network play, etc are all newer features possible by the power of a strong PC with a good internet connection. They can be cycle accurate, but often we need to push the system to its limit (high Ghz CPU etc) to do so.

If your objective is to replace original hardware, or attain a similar "feel" to the original, I would argue FPGAs are better. But we cannot say they are better in every aspect.

MiSTer is actually not just an FPGA (like he Analogue devices), but something beyond that... it combines many benefits of a CPU with FPGA speeds.

Is that "better" for you? That is up to you depending on what you are looking for.

[Newsdee]

For most software emulation means running on a general purpose computer system, with hardware shared between processes.

The main issue I have with software emulation is commercial devices. Most of them are "good enough" setups, with often bad upscaling and lag, running on cheap phone SoC boards barely suited for the task.

So in a commercial context, I trust that anything running on FPGAs will be much better than a software emulation device at a similar price point.

I haven't tried running a SNES on a Steam Deck, but it costs much more than an Analogue Pocket, and either of those will run much better than the average Hyperkin product costing the same as the Pocket.

i can also run Higan on a PC and have a cycle accurate emulator. But you just don't find consumer products using Higan (in ideal setup) out of the box.

[JayRockets]

for how great and accurate Higan and bsnes supposedly were, I was never able to get smooth performance out of either of them. I had been playing snes games on higan, then bsnes, for over ten years thinking those scrolling hitches were how a real snes behaved, haven't having touched a real snes in over 20 years. "This must just be how the hardware behaved, afterall this is just like playing on a real super nintendo, everybody says so, yeah its just slowdown." well when I later discovered it wasn't just "slowdown" and I tried to remedy it I could never get rid of it without either introducing v-tearing or audio glitches.

[Stinky]

for how great and accurate Higan and bsnes supposedly were, I was never able to get smooth performance out of either of them.

That seems to be a common misunderstanding of these emulators which may be deterministically accurate but that does not guarantee they will have accurate latency or performance.

[HerrBerzerk]

Mostly agree for Mister overall but the A1200 example is well outside the bounds of "good enough." Its much faster than the real thing, like a turbo switch you can't shut off, which causes much software to run very differently.

In another thread it is written that the Amiga1200 core is running too fast... somebody know if that is true?

[LamerDeluxe]

Also, don't forget how power-efficient the MiSTer is, compared to the amount of energy needed for a fast enough system to achieve the same accuracy and low latency.

[Armakuni]

Mostly agree for Mister overall but the A1200 example is well outside the bounds of "good enough." Its much faster than the real thing, like a turbo switch you can't shut off, which causes much software to run very differently.

In another thread it is written that the Amiga1200 core is running too fast... somebody know if that is true?

This is due to the TG68k 68020 module not being cycle accurate but does offer great performance due to it.

Some people moan that some games run faster but mostly it's not an issue and we have people shoving PIStorms in their Amigas now to get speeds not ever possible on real hardware before

The 68000 FX68k module is cycle accurate so you can always use that anyway for A500/600

[FPGA64]

The Amiga Core is also not cycle accurate in the Custom Chips, but its good enough that most things run.

[Newsdee]

The PCXT core is cycle accurate in its 4.77 MHz, 7.16 MHz, and 9.54 MHz modes, but is no longer cycle accurate at maximum speed (AT like performance).

[peke]

original hardware vs mister is what I been thinking since I got my mister one year ago....I share my humble opinion....

here´s my setup as today...

https://imageupload.io/0FkfvZEzzE35hkL

I got a lot of stuff I been collecting since years ago....consolized mvs (with 161 in 1 cart), atari 2600 with harmony cart and av mod, all consoles got evedrive/ode and hd retrovision cables, all (except ps2, xbox, wii, 3do and dreamcast....all have cores on mister, which it makes it redundant to have, and where I live this things ain't cheap compared to the salary most earn)

thinking of selling all and keep mister goes over my head since last year...maybe I will do it...but...here´s my opinion:

Mister Is so great...almost perfect, but I don´t know why sometimes feels like soulless compared to the real thing....I can´t explain why....sometimes is an audio issue...sometimes a controller issue,....don´t get me wrong...mister is fantastic and If I have it before I think maybe I would not have all this stuff. Playing 2600 with paddels or nes with a zapper....feels like I use to when I was a kid...when I make a comparison with the real thing side by side....almost always original hardware feels natural and mister is like 99% perfect....but there´s 1% that I can´t explain that keep me away of selling all this stuff....maybe I´m crazy....

I also got a turntable with vinyls and I feel almost the same....I got spotify, mp3, flac and all the music I will (maybe) never hear....but the experience of hear back in black on vinyl is priceless.....It feels like angus is playing his sg with a marshall right to me....

I used to play all the retro consoles on my wii before,....wii got excelent emulators and component output...but still...don´t feel the same...feels like a washed version.

don´t get me wrong....love mister...use it and maybe sooner or later I sell all these stuff!

hope somebody feel like in the same boat! and share his experience!

[thorr]

hope somebody feel like in the same boat! and share his experience!

Using original hardware will always have a special place you can't get anywhere else. For example, the nostalgic look of the Atari 2600 sitting in front of you and looking through your case of cartridges and then inserting one and flipping the switch and possibly getting a glitch and having to re-seat the cartridge and try again, then using the spring loaded select and reset switches and flipping the difficulty switches to see what they do, etc.

I fairly recently got a Nintendo 64 and most of the cartridges I wanted for it. I haven't even played some of the cartridges yet due to other priorities, and now the N64 core is coming. I think for the cartridges I have, I will use my N64 when playing those games. For the others, I will use the MiSTer. It turns out there are a lot more N64 games that I am interested in after watching the videos on Youtube.

As far as the feel goes, this can be absolutely correct depending on the situations. Atari paddles is one of those situations. For some reason, this is an area that seems difficult to perfect. It's 95% good enough now though, but not 100% the same. SNAC could potentially fix this but would still need some tweaking in the programming to make it perfect because of the ADC requirement. I haven't tried it recently, so maybe this is addressed already. Atari joystick games are pretty much perfect though. ao486 is amazing, but another one that is different feeling to me. It is 90% good enough I think. PCXT is much better in this regard IMO, but still has CRT related display issues in some video modes last time I checked that require the scaler to fix. My hope is that both PC cores can exist without the need for the scaler at all someday. Also, a cycle accurate 486 DX/2 66 would be awesome, as well as (if possible) a designed from scratch CPU that ekes out as much performance as possible and adds a Voodoo card for an early Pentium-type performance Voodoo Windows 98 core. That might also not be 100% the same, but I could live with it in order to have access to that era of PC gaming in FPGA.

[Hodor]

original hardware vs mister is what I been thinking since I got my mister one year ago....I share my humble opinion....

here´s my setup as today...

https://imageupload.io/0FkfvZEzzE35hkL

I got a lot of stuff I been collecting since years ago....consolized mvs (with 161 in 1 cart), atari 2600 with harmony cart and av mod, all consoles got evedrive/ode and hd retrovision cables, all (except ps2, xbox, wii, 3do and dreamcast....all have cores on mister, which it makes it redundant to have, and where I live this things ain't cheap compared to the salary most earn)

thinking of selling all and keep mister goes over my head since last year...maybe I will do it...but...here´s my opinion:

Mister Is so great...almost perfect, but I don´t know why sometimes feels like soulless compared to the real thing....I can´t explain why....sometimes is an audio issue...sometimes a controller issue,....don´t get me wrong...mister is fantastic and If I have it before I think maybe I would not have all this stuff. Playing 2600 with paddels or nes with a zapper....feels like I use to when I was a kid...when I make a comparison with the real thing side by side....almost always original hardware feels natural and mister is like 99% perfect....but there´s 1% that I can´t explain that keep me away of selling all this stuff....maybe I´m crazy....

I also got a turntable with vinyls and I feel almost the same....I got spotify, mp3, flac and all the music I will (maybe) never hear....but the experience of hear back in black on vinyl is priceless.....It feels like angus is playing his sg with a marshall right to me....

I used to play all the retro consoles on my wii before,....wii got excelent emulators and component output...but still...don´t feel the same...feels like a washed version.

don´t get me wrong....love mister...use it and maybe sooner or later I sell all these stuff!

hope somebody feel like in the same boat! and share his experience!

I own several 8 bit and 16 bit computers and I will never replace them by any FPGA. Sometimes I tinker, play or simply watch a demo with the equivalent core, but the experience and feeling that the original hardware offers me is absolutely irreplaceable.

[peke]

music is another pasion I have as a hobby...I play the bass

in an interview once billy sheehan say he doesn't use active pickups beacause when there's a sound issue he always ask himself....is it the battery?

maybe I feel the same about emulation...love it but...when something is not like I use to remember i ask to myself...is the emulation?

I know it's me and maybe for 99% of the people emulation is fine...

I use mister for the stuff I don't own and for convenience some times (maybe my kid wants to play nes and he isn't a perfectionist.

[PikWik]

in my opinion, the MiSTer is a system that appeals to some niche audiences.

1) speedrunners of retro games who want an all-in-one FPGA system for retro consoles. added benefit of HDMI output for streaming and video game capture. these group of users are fans of the MiSTers dual video output for streaming/playing, no input latency, and accurate audio/video sync.

2) people who want an experience of retro consoles with as little fuss as possible, with gameplay being as close to authentic as possible, and to know the FPGA is presenting the console or computer in a correct timing and/or cycle accurate. input latency not being a top priority and introducing lag in the form of screen rotating, bluetooth controllers, and BFI is not an issue.

3) arcade cabinet owners who have existing JAMMA gear who want a "better" way to connect their cabinets to a device that can play arcade games from. groovymame is nice and is probably all most people would need. but if you are wanting a cutting edge way to play old arcade games, the MiSTer is it.

4) general enthusiasts with disposable income. people who want to see what everyone is talking about and have tried out the current state of software emulation on their devices they already have (retroarch, rasp pi, mini consoles).

---

ive been emulating these systems since the early 2000s and have used a bunch of them over the years.
thinking about when they have impressed me the most, there have been a couple standout moments. first was the SNES emulator, ZSNES, and finding all of the RPGs i had missed out on growing up. getting into those 2D pixel art masterpieces was so memorable. another was getting into MAME and playing all of the arcade games that cost me real money to play in the 90s. but now i could press a button to insert a coin

the MiSTer has absolutely been another. i was honestly just fine with the current state of emulation before FPGA. i had used all of the emulators up until this point, and was content with their performance. it wasnt until getting into some very input sensitive games like kaizo super mario romhacks that i started paying attention to how the best people were playing the SNES. it seemed like FPGA was their secret. i bought the MiSTer as a cheaper super NT replacement at the time, with the added bonus of whatever else it could play.

but wow, FPGA really opened the doors of retro gaming. i think its something about only having one version of a core to use, which has a liberating feeling not having to choose between so many options for one system. that limitation forced me to not worry about the core and just focus on quality games for each system. when its so easy to play each system/computer, i can enjoy my time diving into a library to find hidden gems.

and while its not a guarantee every core will be cycle or timing accurate, FPGA is top quality retro gaming, no question

[Bristles]

but the experience and feeling that the original hardware offers me is absolutely irreplaceable.

Same for me. One thing Mister can never have is history, because it's a relatively new device, and only a few years old. Yet, if I get one of my retro consoles out of it's box, the Atari 2600 for example, it has history. And this history would have included times where me, and my family, are sat hudled around a small 14" colour TV - that is all we could afford back in 1977/78/79, and we would take turns playing Space Invaders. Or, coming home from shopping with my Dad, having just bought a new 2600 cart, I would be dying to try this cart out, but had to eat my dinner, wash up the dishes, then connect-up the 2600 ready. Great days, and I still have the same 2600, same controllers, carts, and this brings back wonderful memories.
The same for my ZX Spectrum, still working, but now with composite mod, DivMMC SDCard solution, and Maxduino digital tape unit. Nothing beats using the real Spectrum, because you know it's the same machine that rolled off the production line back in 1983/84, built here in the UK. Although it's a nice touch having a Recreated ZX Spectrum keyboard on Mister, albeit a slightly different rubber feel to the real thing, it's still close.
Memories, and nostalgia can be strong, and holding the very same hardware that gave you all those wonderful memories can never be beaten.
The Mister, to me, is simply for convenience, as there are times when I just want something ready to go, without the faff of connecting up the real thing.
And space is a reason for Mister, since my real hardware would fill every room in my home if I connected it all up, and used CRTs. But Mister connected to a CRT still makes the hair on my arms stand up, because, especially at night, you can almost forget you are in-front of a tiny box, not real hardware.

For emulation, I just stick to newer 3D systems such as PS2/PS3, Dreamcast, Xbox, Xbox 360, Wii/Switch etc... You can do some amazing things with modern emulation, such as 4k upscaling, texture packs, 30fps to 60fps patching. Just look at that Atari Jaguar emulator by Rich Whitehouse, it's superb, and possibly even better than the real thing, thanks to some amazing enhancements by Rich - including some amazing Alien vs Predator enhancements.
What I am thankful for, no matter what you prefer, emulation, real hardware, Mister, is that we can still play all these amazing retro games, and they're not lost to time.

[Armakuni]

The Amiga Core is also not cycle accurate in the Custom Chips, but its good enough that most things run.

I knew about the CPU but not the chipset, as you say it's good enough

The C64 core is still not 100% and that is even older than Minimig in origins but is good enough since the work Sorg put into it a while ago.

There are more accurate closed core products if you want to focus on a single system too