Intro

So… It’s been a while since I last updated my blog. It was never a “job” for me, I just wrote when I wanted to say something that I can easily recall in the future and share. I do need to make an update on the last subject I posted (about Retro Collections), since many things happened since… exactly 2 years ago (oups!), but staying on the main subject of “retro”, I will write some things about the current state of this branch of re-living older machines, that is called FPGA.

For people that may not know the concept, a very rough sketch of what this is, with crude generalizations that some may bother some, but I will do it anyway:
So FPGA is a kind of integrated circuit, that doesn’t work like a general processor (CPU), but rather emulate the actual circuits that comprise the “normal” chips. This way you can make the FPGA “be” a specific “CPU” and/or “GPU” and/or other custom chip (depending on the “capacity” of the FPGA). Normally FPGA was created exactly to simplify testing of chips under development, but emulation community saw the opportunity of using those chips to instead work as existing older hardware. Current “nicely priced” FPGA are able to “dress” like most 8bit hardware we know, which depending on documentation we have on that hardware, may be astonishingly accurate, “like the real machine”.
FPGA fans will attack you if you talk about FPGA like “emulation” as they (I do it to) fight that FPGA is NOT emulation, but actually the closest thing to the real thing. This of course in the end has to do with the actual implementation, for example some machines’ custom chips are not well documented, but rather “reverse engineered”, so in practice some parts MAY be in actuality “emulated”. These of course are semantics to most people and in the end, to the user, the idea is that FPGA in general is “closer to the real bare metal than emulation can be”.

With the above explanation aside (and hooray to me for not using AI to write it), let see the state of things as is in  late 2026. Note that I am clearly biased towards Commodore family and may have ignored FPGA developments, towards other retro machines. Some are ignored on purpose as they are “too small”. Feel free to add extra info.

 The systems

MiSTer

This probably will be 80% of the article, as it is “MAME” (or for some the “RetroArch”) of FPGA, from which many other projects stemmed, it is the most generic out there, with super active development.

So, MiSTer is an open source project, that was born as the successor to MiST (which btw still gets some hectic development), which as the name suggests, started as an ST 🤮(sorry) and Amiga (“Mi”) FPGA project. It made many improvements on the board and sub-boards and used a stronger FPGA, DE10-Nano, that had the cool characteristic, that was sold as a ready made dev board with special pricing for students and includes basic I/O and even a (not very strong) ARM that can be useful to things that are not to be “within” the FPGA. Later, things became a bit more expensive (it is not that DE10-Nano boards grow on trees), but Chinese knock-offs became available and even some serious DE10-Nano based “re-implementations” we’ll discuss later.

So, MiSTer is sold as a modular system, where you have the base dev board, you get sub-boards (like extra I/O, USB, RAM, audio etc.) that plug on it and some nice (or not so nice) boxes to pack them all together. Then you get the software and cores (core is the actual software module that “teaches” the FPGA what to do – what machine to be), that nowadays is a very streamlined operation, easy to follow and keep updated.

This is where the fun begins. MiSTer’s wealth of cores, beats the hell out of other FPGA projects. Actually other FPGA projects, just port MiSTer cores to similar or different FPGA chips (after all FPGA programming languages AFAIK are exactly two – maybe two and a half). You will find most, if not all 8bit systems you remember, many many 16bit systems also (be it computers, consoles or whatever) and even many arcade cores! There are at least a few core developers and as I said the project is VERY active (even though it is a few years old now) and new breakthrough is still happening, like hybrid cores (cores that use FPGA for what matters and ARM CPU for what doesn’t need that kind of handling). I will not go into details, this article is not to give all the details but mostly to trigger you to look up the extra details yourself.

Just some of the capabilities of MiSTer are:

• Ability to use virtually zero latency controller connectors (something Emulation can never ever touch), that you can swap and connect your favorite retro joystick (not just DB9).
• Ability to display both to digital (HDMI) and analog (VGA/RGB/Composite), with the analog side again being virtually latency free.
• Ability to have tricks like NFC triggering. In other words you have a “Pac-Man” NFC card, you touch the system with it and in a few seconds, you get arcade Pac-Man. Or some other game, or a machine. It is like a modern “cartridge” feeling.

So, now there are 3-4 retro-shops worldwide that will sell you “full” systems (i.e. include all the modules you need, even pre-assembled) and as I said even some re-implementations I will refer to below:

MiSTer Pi

This is just an alternative to the main DE10-Nano board, in a much cheaper price (no relation to actual Pi hardware, which -Pi- had nothing to do with FPGA). Hard to find nowadays as its creator Taki Udon, now works mostly on…

SuperStation One

This is a custom DE10-Nano system, fully compatible with MiSTer, but is All-In-One and PS1 themed, with PS1 controller ports (which does make it a no-no for me, but this is just me, after all it is a “normal” MiSTer). A very good option for people that want MiSTer but want less bother and a more polished system.

MultiSystem2

This is also an All-In-One alternative DE10-Nano system, that is fully compatible with MiSTer. To me is a more attractive project than Taki Udon’s, but also possibly a more “risky” one. It comes from a UK based company, but the box is 3D printed (which at this point I would prefer to print myself but they don’t give me the option) and still has extra modules to add to make it (more) complete, that may rise the price bar, but does give you everything you may need in a more console like box, that is not PS1 themed (I say this as a plus). I am a bit scared of how strongly they can support their system though – possibly for no reason, just a feeling (sorry).

Of course with ALL these systems, I probably need to make a disclaimer for them, you understand that you do get a system “as is”, with the limitations they have and without huge companies backing them up.

MiSTer successor

This is purely theoretical right now. Eventually as MiSTer was born from MiST, a stronger “MiSTer 2” has to evolve. FPGA experts though, say that this cannot be a revolution right now (i.e. suddenly be able to build a PS4), as current FPGA resources are mostly OK for all 8/16bit systems except some late ones that are more complex and POSSIBLY even moderately stronger FPGA will still not be able to implement. But I believe a MiSTer successor WILL eventually come.

Apollo/Vampire

This is an Amiga based project I initially didn’t see with much sympathy, as I thought the team that is involved in it, was trying to split the Amiga world (and I am very biased with everything Commodore and Amiga), capitalize on something it was not their to capitalize (mostly because of their AmigaOS build and selling it).

I am less aggressive nowadays and have to admit the serious work they have put on this, that is not just a “retro” Amiga project, but also an attempt to EVOLVE the Amiga.
What do I mean. Well, instead of just implementing known 68000 variants, and known Amiga chipset variants (OCS/ECS/AGA), they also made an imaginary 68080 CPU (trivia: there is an 68080 in reality but is NOT a CPU and not successor to 68000 series), that is much stronger than 68060 (last real 68K CPU), with all known 68K series instructions and many extra ones and also “SAGA” chipset that is a much stronger successor to AGA, with included RTG – something that Commodore themselves didn’t have time to really evolve. It is definitely interesting and as a team (and company) they seem to endure over the years, evolving the original project.
Add to that (for those interested because they had the wrong machine back then), that ST core also gets advantage of SAGA and virtual 68080, so they actually evolved ST line also.

Right now Vampire exists as an add-on to A500, A1200, A600, A2000 even A1000, but for some reason I don’t see A4000 mentioned (WTF… I have an A4000) AND ALSO as a stand-alone computer with embedded keyboard, that looks like a modern A600 (would prefer a modern A1200 shape as I want to have num keyboard) and another stand-alone model that is just a small box you can plug peripherals on.

I have to say the extra features and faster response, is something I would like to… have the option to use (read: play with). It is nice to play that old Amiga game you remember, but if you did own the computer, you also would love to see what machine never happened and could happen.

Ultimate64/Commodore 64 Ultimate

So this is interesting. Once upon a time, a guy called Gideon, made an FPGA board, that could replace your broken down C64, in the original case, with even some extra features (like freezer actions and multi-SID support). This project evolved with a couple of newer versions. This was already very interesting, as also breadbox and C64C cases, were made and people could in the end have an almost real NON-emulated C64, using modern hardware and some modern fascilities.

Years after this project already existed, there was a guy (I will make this VERY brief, I don’t want to divert the article), that dreamed to wake up Commodore, get the patents, the name, get even some of the still-on-this-planet people involved in original Commodore. He recently DID succeed to a point (this deserves a whole article by itself, but you will find some if you search), so now we DO have an “official” Commodore once more, although we know now that they didn’t get ALL we hoped, some things are under license and Amiga is out of reach still (hello Mike, I see some action by you recently – the keyword man, is “co-operation” – please consider). The decisions new Commodore take are not always to my liking though. For example they started by pushing down to our throat a supposedly “new Commodore” machine that is (as they still sell it) in reality a mini-PC in a breadbin case, with a Linux distro that is skinned in a very weird “early-2K web page” way (flashing colors all the things we want to forget), that gets tiring very soon and has no-relation to feeling like Commodore (except using Commodore colors, fonts and such) and in reality you get “stock” CBM and Amiga emulation (not even complete, they don’t even include the roms!), via VICE and fs-uae (in turn based on WinUAE), things that people that want to emulate CBM machines and Amiga machines already use and not need get a weird breadbox PC. I have more examples of what I believe is money-grab policy, but not a matter of this article.

But another project of theirs, was actually interesting. They actually collaborated with Gideon and made some further evolution to Ultimate64-II and sell it as a complete “new” C64 “Ultimate”, which in many ways accurately describes the machine, as it is an FPGA C64, with many extra features, we would love to have back in the 80ies anyway. The price is also not super expensive.

So yes, an interesting project, but then you are struck with the (forced if you ask me) limitations. First of all this is “just” a C64 (even Ultimate). This is the only core in there and they definitely could (and are even legally allowed AFAIK), include all other CBM 8bit machine cores (if I have to remind you, PET, VIC-20, +4, 128 etc.). I even tried to talk them into doing it (to the point of being “that weirdo”). Also they could maybe implement (actually port) other 8bit cores or even some 16bit cores (Amiga, hello?) to make the machine even more attractive. And scratch this, not even include those cores BUT allow people to run other cores that 3^rd party would port. Why on Earth NOT allow this? The FPGA in the machine can do it. They actually just didn’t allow for it. So although I loved the project (as Ultimate64-II before it), I have many second thought and for now I am not touching it.

Analog (Pocket)

This is an interesting company, that originally came up with a very nice idea. Implement FPGA in handheld format! So they did!

They made a very nice handheld, originally (officially) made to run well known handheld machines (and be modular enough to accept their catridges directly!),  but also other cores got implemented, so you can for example have a “handheld C64” (although I think the core is experimental). Definitely an one-of-one FPGA project that deserves mention.

They also made some brand-specific non-pocket FPGA consoles. I won’t talk about them, because they are on purpose too specific (for example Sega based and Nintendo based) and AFAIK with no extra cores and you already know from the rest of the article that this is something I do not love.

Mega65

I will start with the elephant in the room. Mega65 is a very expensive project. Even now that people start to like it more (more on this below), it is still difficult for most to just get it. It is almost 1K to buy it. BUT, is a very good quality product made by a small team (not a big company, even if the company is a startup like new Commodore) and to get to you this kind of quality in such small quantities, you need to pay for it.

But what is it? A tiny bit of history first. Some time after C64 (and its immense success almost worldwide) and after Amiga sub-brand was already established (at the time of A500 and before A1200), they saw a gap between C64 and Amiga 500 (and maybe they were right, if the price was right – although I have to say that similar projects like Amstrad CPC6128+ were failures). Something compatible with the HUGE library of beloved C64 (on hardware level, UNLIKE the Amiga), but way stronger and with more features (and yes, way better BASIC). We heard about this machine much later, as it never got released and just around 200 early dev devices got spread around. They are VERY expensive collectible items now. C65 would be a nice all-in-one with a 3.5” drive, facing front not side (i.e. like CPC6128).

Years later and after some of the early firmware was already found and immortalized, a small team, thought of actually implementing something close to C65, in FPGA. The result was Mega65. Mega65 is actually earlier project than many of the others mentioned here, but was always a niche product, that recently got a new look by many people (including yours truly), after the release of C64U by new Commodore (see above).

Why? Because it is what C64U COULD (or better SHOULD) be, it is a bit what Vampire is but for 8bit CBM, but with the openness (but not spread) of MiSTer.

You get a C65 shaped machine (i.e. something looking nice and very “Commodore like”), not 3D printed or anything, that implements NOT only C65 (from the available hardware specs we now know and KERNAL ROMs that exist), but also other Commodore AND non Commodore cores. You also get a C64, an (early) C128, +4, VIC20, Spectrum, QL and few others. More over (this is where the Vampire parallel comes), it comes with Mega65 core. In other words a machine that COULD be C65 or even an evolution of C65, as it has features and strength close to many 16bit systems. It comes with a very nice like original C64 printed manual (who didn’t love that manual and in general the complete manuals of the era?) and box. But you get financial pain for it. Again, it is what C64U could be (and would also be cheaper than Mega65 as the scale and scope would be different), which is why people start seeing it again, comparing and discussing it.

Misc FPGA

Well not an exhaustive list, but those projects are too small to get their own section and even smaller than THESE, I don’t even mention.

• ZX-Uno, a nice Spectrum ZX based FPGA, that also has some other 8bit cores implemented. Not sure if you can buy it currently, as it is mostly of the MiST generation.
• ZX Spectrum Next, another one, not very active currently (was mostly a kickstarter project), with very few extra cores.
• Suska, Atari ST (and even Falcon) based project. I don’t know much about it, I think mostly an one guy project. Not very active.
• Game Bub, potentially a game changer, as it will be the only competition to Analog Pocket, and also be Open Source! Right now only Gameboy family (and cartridge slot) is implemented, but when more cores are developed/ported… who knows. I think it does not have the cartridge slot modularity of Analog Pocket though. Also has stronger FPGA but less screen resolution. We’ll see.

Etc.

Upcoming Projects

I will mention just one actually, as is the only one I see serious enough development actually happening. I am waiting for it, VERY eagerly. It is called Replay2.

Replay2 implies there was a Replay project and indeed there was. It was an FPGA board, Amiga based (they even implemented a physical 68060 slot on it), supporting cores, but in reality it never got many sales, it never got much development, it never got many cores (there are a few).

Why I am eager about Replay2 then? Well there are a few reasons…

• Although the team’s marketing plainly SUCKS (they really look dead to people not really looking deep into the project – I have asked them to improve this very important thing called marketing more than once), the actual development is very active and we are very close to a real new board.
• Since now MiSTer exists, actual cores will be easier to find and eventually port. It is a different playground than the era of Replay1.
• It is the only actual “next gen FPGA” project I can find, not theoretical, but actually almost ready to build. Potentially it CAN be the next MiSTer (in a less modular way, but that already implements most that you would add anyway), if the team (and resellers) handle it right.

The only half decent source of info about Replay2 (in their site you can find only some ghost small article), is their discord and even there, you will have a hard time actually get “user friendly” data on the status, features etc. It is mostly like the discord is from dev board devs to dev board devs. But you can see work done.

Aside from Replay2, we may hear about MiSTer with stronger (compatible) FPGA in the next few months and who knows what else.

The makers of MultiSystem2 are (early) in the development of a pocket version, with modularity being the stepping stone (they intend to make it usable for consoles, arcade, computer cores). This would be the 3rd portable FPGA player, but it is still early.

Other projects have popped from time to time (like the disaster called MARS), but I do not see other really active projects big enough to attract attention right now.

Closing…

This is a very exciting field of retro “emulation” (ok not emulation, but you know what I mean), that has two important factors that may trigger you:

1. You get something with a strong family-acceptance-factor (FAF, used to be WAF), because it is something that can be as plug’n’play as a console of yesteryear. Nothing to “boot”, nothing to go wrong. If it works, it works.
2. The way you come “close” to the real machine implemented, cannot be achieved even with modern hardware running modern emulation. It is something you can “feel”, the no-latency of the real machines, esp. if running on a CRT monitor.

I for one eagerly wait future development in the field. Here is my dream machine, that will probably never come:

Something in a “pizza box” very similar to Commodore 128D or Amiga 1000 (machines that I dream of getting and getting to display them), with the flexibility of MiSTer, but actually a stronger FPGA like the ones that are already upcoming (Replay2 board?), an actual CPU (a multi-core modern ARM?) on the side (also for emulation) and plenty of cores (all 8bit and 16bit computers/consoles/handheld/some arcade), even some weird ones like C65, Mega65 and Vampire V4. Esp. for the last part, it would involve many people to agree together, something that in my (allow me to say vast) experience in retro-computing has taught me, it is VERY rare and many times even starting nicely, results in conflict. For some reason such a small community instead of uniting to become stronger, tends to split up.
With that (negative) comment, I close this article.

Until next time.