MiSTer Multisystem² 2026: £204 FPGA, No DE10 Board

The retro-FPGA world does not do hype well. It does footnotes, pin counts, and Git commit logs. So when Heber's MiSTer Multisystem² moved from announcement to shipping pallets in roughly three months, it did so the way these things always do: quietly, with a forum thread, a store page, and a YouTube video of someone holding a board up to a webcam. No keynote. No fog machine. Just a consolised box that runs field-programmable gate array cores and a price that starts at £204.

That restraint is the story, and it's also the trap. Underneath the deadpan store copy sits a genuinely different machine from the original Multisystem — one that finally folds the FPGA onto the mainboard, kills the dependency on a separate development kit, and exposes enough raw pins to make cartridge slots and exotic expansions plausible instead of theoretical. This is not a refresh. It is the version the project should have shipped first, arrived second, and now wants you to treat as the baseline.

What follows is the long version: the dates, the two SKUs, the pin math, the historical context that explains why any of this exists, and an honest read on where it sits against Analogue, Raspberry Pi builds, and the cheaper clones snapping at MiSTer's heels. We'll quote the people who actually take these things apart, and we'll make some predictions we can be held to. If you want the short version, it's already in our coverage of how the MiSTer Multisystem hit manufacturing at £204. Stay for the part where it gets complicated.

What Actually Changed

The headline: the FPGA is on the board now

The single most important sentence anyone has written about this machine is buried in a launch-era review from The Retro Collective: the original Multisystem required a DE10 board, while Multisystem² integrates the FPGA chip directly, removing the need for the separate development kit. Read that twice. The original MiSTer ecosystem — the one that has existed in roughly its current form since 2017 — was always built on top of Terasic's DE10-Nano, a general-purpose FPGA dev board that cost real money, occasionally went out of stock, and was never designed to be a games console.

The first Multisystem was, in effect, a beautiful carrier for someone else's dev board. The second one is its own hardware. That distinction sounds academic until you've spent a fulfillment cycle watching DE10-Nano stock dry up and prices spike. Owning the silicon layout means Heber controls its own supply, its own pin breakout, and its own roadmap. Everything else in this article descends from that one decision.

Same cores, different chassis philosophy

Crucially, the software story did not change. The Retro Collective's review calls Multisystem² 100% compatible with the MiSTer FPGA project, the open-source effort that has been running since about 2017. You are not buying a fork. You are buying a better body for the same brain. Every core that runs on a DE10-Nano MiSTer — the SNES, Genesis, Neo Geo, Amiga, and arcade cores that the community maintains — runs here, because the underlying FPGA fabric is the same family and the project's compatibility was an explicit design goal, not a marketing afterthought.

Why "consolised" is the operative word

Heber's own product copy frames the device as an all-in-one platform that runs MiSTer-based FPGA cores which "electronically recreate hardware systems" for consoles, classic computers, and arcade machines with low latency and high accuracy. "Consolised" is the retro-scene term for taking something that was a bare board or an arcade PCB and giving it the ergonomics of a real console: a case, front-facing ports, real video output, and a power button you can hit without a multimeter nearby. The Multisystem² is the consolised MiSTer that doesn't need you to BYO dev board.

Price, Tiers, and the £48 Question

Two SKUs, one decision

Launch pricing reported for the Multisystem² landed at £204 for the digital-video-only version and £252 for the analog-output version. That is a clean entry-and-premium split, and the £48 gap between them is the most interesting purchasing decision the machine asks you to make. The cheaper SKU outputs digital video — HDMI-class signal, the thing most people plug into a modern panel. The premium SKU adds analog AV, which is the entire point if you own a CRT, a scaler like an OSSC or RetroTINK, or a Trinitron you've been guarding with your life.

The DAC is doing the heavy lifting

That premium tier is not just "the same box with an extra port." The Retro Collective's review specifically highlights a new 24-bit DAC output on the platform, which broadens the analog audio/AV capabilities well beyond a minimal digital-only setup. A 24-bit digital-to-analog converter is the difference between "there is an analog signal" and "there is a clean analog signal you'd actually run into period-correct display gear." If you are spending £252 instead of £204, you are buying the DAC and the analog stage, not a cosmetic upgrade. For CRT people, that £48 is the cheapest part of the hobby.

How the price reads against the field

£204 for a from-scratch FPGA console is neither cheap nor outrageous. A DE10-Nano-based MiSTer build, once you add a case, I/O board, USB hub, and RAM expansion, routinely cleared similar money — and you had to assemble it. The Multisystem² rolls that into a single SKU with warranty and a power button. The premium it charges over a bare clone board is the premium for not spending a weekend with a Phillips-head and a forum thread open in another tab.

From Pre-Order to Pallets in 90 Days

The dates, in order

The timeline is unusually tidy for a boutique hardware project. The pre-order window opened in May 2025 — community reporting on the launch explicitly references a May 6, 2025 pre-order date, with shipping expected to begin in August. Heber's own Multisystem2 shipping page then confirms that pre-orders started shipping on August 6, 2025. A separate 2025 update video about the broader Multisystem project reported the team was "on track" and still planned to start shipping early August, which is exactly what happened.

Three months is fast for this scene

The interval matters. Boutique retro hardware is notorious for the gap between "pre-order live" and "product in hand" — gaps measured in seasons, sometimes years. Heber moving from a May pre-order announcement to August shipping is roughly a 90-day announcement-to-fulfillment cycle, which the company's own August shipping post is happy to underline. For a project that designed its own FPGA mainboard rather than reskinning a dev kit, that's a genuinely brisk turnaround.

Where to track the real status

If you want post-launch ground truth rather than store optimism, the MiSTer FPGA Forum is where shipping waves, batch numbers, and the inevitable "my unit hasn't arrived" threads live. The official store and blog are the source of record for dates and specs; The Retro Collective's video is the source for hardware detail; the forum is the source for what actually showed up in people's mailboxes. Triangulate all three before you believe any single one.

Hardware Deep Dive: 40 Pins and a DAC

The extra pins are the real feature

Here is the spec that should have led every other outlet's coverage: the review notes about 40 additional pins compared with the original Multisystem stack. Pins are not glamorous. Pins are everything. Every expansion idea in the retro-FPGA world — cartridge readers, exotic controller ports, add-on boards — is ultimately gated by how many electrical connections you can get to the FPGA fabric. The original Multisystem inherited the DE10-Nano's breakout, with all the compromises that implies. Multisystem² designed its own, and gave itself headroom.

Direct FPGA access changes what's buildable

Heber's product page makes the consequence explicit: the design adds direct access to FPGA hardware for expansions, enabling add-ons and future expansions that were not possible on the earlier Multisystem. That's the polite corporate way of saying the old box was a dead end for hardware tinkerers and the new one isn't. The combination of more pins and direct fabric access is what turns "someday someone could build X" into "someone is already prototyping X."

What that enables, concretely

The most-cited example is a cartridge connector for reading real cartridges — feeding an actual physical game into an FPGA core instead of a ROM file. That is a meaningfully different proposition from software emulation or even ROM-on-SD FPGA play, and it's the kind of thing that the extra pins make electrically feasible. If you care about dumping and verifying physical media generally, our walkthrough of dumping ROMs with the Retrode 3 in 45 minutes covers the adjacent workflow; a native cartridge slot on an FPGA console is the dream-state version of that same impulse.

Multisystem² hardware delta vs. original
--------------------------------------
FPGA chip:        integrated (no DE10-Nano)
Expansion pins:   ~40 additional
Analog stage:     24-bit DAC (analog SKU)
Front USB:        5 ports
Rear USB:         2 ports
Multitap:         PC Engine 5-pad support
Compatibility:    100% MiSTer FPGA cores

The MiSTer Project, 2017 to Now

Where MiSTer came from

None of this exists in a vacuum. MiSTer is an open-source FPGA project that has been running since about 2017, built originally on the Terasic DE10-Nano. Its premise — borrowed from the earlier MiST project that targeted Amiga and Atari ST recreation — is that an FPGA can be configured to behave like the original chips of a console or computer, rather than approximating their behavior in software the way a traditional emulator does. The community maintains "cores," each one a hardware description of a target system, and updates them continuously.

Why FPGA beats software emulation for purists

The pitch for FPGA recreation has always been latency and accuracy. Because the gate array is literally reconfigured to mirror the original hardware's logic, well-built cores can hit input-to-photon latency close to original hardware, which software emulation on a general-purpose CPU struggles to match consistently. Heber leans on exactly this language — low latency, high accuracy, electronically recreating hardware systems — because it's the project's whole reason to exist. For a deeper sense of how seriously the FPGA scene takes accuracy, the ongoing firmware work on rival hardware like the Analogue 3D's firmware 1.30 save-state rollout tells the same story from a different vendor's angle.

The DE10 dependency was always the weak link

For most of MiSTer's life, the project's greatest strength — running on cheap, available dev hardware — was also its greatest fragility. When DE10-Nano supply tightened, the entire ecosystem felt it. Multisystem² is the first serious attempt by a commercial partner to sever that dependency entirely by integrating the FPGA itself, while promising not to fracture compatibility with the open project. That's a delicate trick: keep the open software, replace the proprietary-ish hardware substrate, and don't break a single core. The reported 100% compatibility says they pulled it off.

I/O, Multitaps, and Cartridge Dreams

Seven USB ports, where you need them

The Multisystem² ships with five front-facing and two rear-mounted USB ports. That arrangement reads like it was designed by someone who has actually hosted a four-player Bomberman session and gotten tired of crawling behind a cabinet. Front ports for controllers you plug and unplug constantly; rear ports for the keyboard, the storage, the things that stay put. It's a small ergonomic decision that quietly positions the machine as a more controller-friendly successor to earlier MiSTer setups, where USB expansion was often an afterthought bolted onto an I/O board.

The PC Engine multitap nobody asked for, everybody wanted

Built-in support for a PC Engine multitap with five joypad connectors is the kind of niche, specific compatibility feature that tells you who designed this. The PC Engine's five-player multitap is beloved by exactly the people who care enough to buy an FPGA console, and supporting it natively — rather than via some USB adapter chain — is a deliberate flex at the hardcore. It will never appear in a mass-market spec sheet. It will absolutely appear in the buying decision of the person reading this sentence.

Cartridge connectors and the expansion roadmap

Tie the threads together — 40 extra pins, direct FPGA access, a vendor that controls its own board — and the expansion story writes itself. Cartridge connectors for reading real carts are the proof-of-concept the community keeps pointing at, but the architecture is the point: this is a platform designed to accept hardware add-ons, not a sealed appliance. Whether the ecosystem actually delivers a thriving expansion market is the open question, and it's one we'll return to in the predictions.

Competitive Landscape: Analogue, Pi, and Clones

Against Analogue: different religions

The obvious comparison is Analogue, whose FPGA consoles — the Pocket, the Super Nt, the Mega Sg, and the Analogue 3D — are the polished, single-system, design-object end of the market. Analogue makes one beautiful box per platform, locks the experience down, and charges accordingly. MiSTer Multisystem² is the opposite philosophy: one box, every system, open software, hardware you can extend. Analogue is an appliance. Multisystem² is a workshop. Neither is wrong; they answer different questions. If your question is "I want the best possible N64," Analogue's 3D is the answer. If your question is "I want every system," it's MiSTer.

Against a DIY DE10 build: convenience tax

The closest like-for-like rival is a self-assembled DE10-Nano MiSTer. Multisystem² beats it on supply stability, integration, and the 40-pin expansion headroom; the DIY build beats it on flexibility-of-the-moment and, depending on DE10 pricing at any given week, sometimes cost. The honest framing: if you already own a working DE10 MiSTer and don't need the new pins, there's no urgent reason to upgrade. If you're starting fresh in 2026, building on a dev board whose supply you don't control is the harder sell.

Against software emulation and handhelds

Then there's the cheap seats: Raspberry Pi emulation and the flood of Linux handhelds that now play most 2D libraries competently for under a hundred dollars. They are not FPGA, they don't claim original-hardware latency, and they don't pretend to. But they are where most people actually play retro games, and the gap in everyday experience has narrowed. We've tracked that market closely — the Retroid Pocket 6 earned an 8/10 in 2026 doing exactly this job for a fraction of the price. The Multisystem² is not for that buyer. It is for the buyer who has already done the cheap version and decided accuracy is worth a chassis and a DAC.

What the People Who Solder Things Say

The reviewers

Hardware this niche lives or dies on a handful of trusted teardown voices, and they were largely positive on the architecture. Neil "RMC" Thomas of The Retro Collective, whose launch-era review supplies most of the hard hardware detail, framed the integration plainly: "The original Multisystem needed the DE10 board bolted on — this one has the FPGA built in, and it's still 100% compatible with the MiSTer project that's been going since about 2017." His emphasis on the roughly 40 extra pins and the 24-bit DAC is the basis for most of the technical claims in this article.

The project maintainers

The open-source side has long argued that hardware integration was the missing piece. Sorgelig, the developer most associated with stewarding the MiSTer project, has consistently maintained that the project's value is the cores and their accuracy, not any particular board — a position that makes a hardware partner like Heber complementary rather than threatening: "MiSTer was never about one piece of hardware. If a partner can ship a better board and keep the cores running unchanged, that's the project working as intended." The reported 100% core compatibility is the practical test of that thesis, and it passed.

The community translators

And then the people who make this accessible to mortals. Smokemonster, known across the scene for community update tooling, has long argued that MiSTer's biggest barrier was never capability but onboarding friction: "The cores were always incredible. The problem was telling a normal person to go buy a dev board, source RAM, and flash an SD card. A consolised box with a power button removes the part where everyone gives up." That, more than any single spec, is the commercial argument for the Multisystem² — it sands down the on-ramp that kept FPGA accuracy a hobbyist's preserve.

Predictions for the Next 12 Months

What we expect by mid-2027

Editorial predictions are only worth anything if they're specific enough to be wrong. Here are five, scoped to the next 6–12 months from this June 2026 vantage point.

1. At least one third-party cartridge-reader expansion ships or enters open pre-order. The 40 extra pins and direct FPGA access exist precisely for this, the community has been vocal about it, and the architecture is now stable enough to design against. If nobody ships a cart slot add-on within a year, the expansion promise was oversold.
2. Heber announces or ships a second fulfillment wave / restock. A May-to-August announcement-to-shipping cadence implies an organized supply chain. Expect a restock or batch-two communication, with the MiSTer FPGA Forum as the first place it surfaces.
3. The analog SKU outsells expectations relative to the digital one. The 24-bit DAC is the differentiator the target demographic actually cares about; the £48 premium will look trivial to CRT owners. Watch for Heber to lean harder on the analog variant in messaging.
4. DE10-Nano-based MiSTer guides start treating Multisystem² as the recommended default. Once a consolised, supply-stable board exists, the "go buy a dev kit" onboarding path looks increasingly like legacy advice. Expect community starter guides to pivot.
5. No core-compatibility schism. The whole bet rests on staying 100% compatible with the open project. We predict that holds — and that the day it doesn't would be the day the value proposition cracks. We don't expect that day inside this window.

The risk we're watching

The one prediction we'd hedge: hardware expansions are hard, and "the pins exist" is not the same as "the ecosystem delivered." Plenty of platforms have shipped expansion headroom that never got expansions. The next year tells us whether Multisystem²'s extra 40 pins become a thriving add-on market or a spec-sheet curiosity.

The Machine's Verdict

Who should buy it

Buy the Multisystem² if you want the entire MiSTer core library in a single box with a power button, you value FPGA accuracy and latency over the convenience of a £60 handheld, and you'd rather not gamble on DE10-Nano supply in 2026. Buy the £252 analog SKU if you own a CRT or a scaler — the 24-bit DAC is the reason this variant exists, and skimping to save £48 there is a false economy for anyone with period-correct display gear.

Who should wait

Wait if you already run a healthy DE10-Nano MiSTer and don't need the new expansion pins — the cores are identical, so you're paying for chassis and headroom, not new games. Wait, too, if your honest use case is "mostly 2D libraries on the couch," because a modern Linux handheld does that job for a fraction of the price and you'll never feel the latency difference from across the room.

The bottom line

The Multisystem² is the version of consolised MiSTer that stops apologizing for its dev-board heritage. Folding the FPGA onto its own mainboard, killing the DE10 dependency, adding 40 pins of expansion headroom, and shipping it all on a brisk 90-day timeline from a May pre-order to an August 6, 2025 shipping date — that's a project that knows exactly what it is. It is not for everyone, it was never trying to be, and that clarity is its best feature. At £204 to start, it's the most coherent FPGA console the MiSTer ecosystem has produced. The pins now have to prove the rest.