MiSTer Multisystem 2: £204 FPGA Console for 2026

For roughly a decade, the dirty secret of the MiSTer Project was that the most accurate retro-gaming hardware on the planet looked like it had been assembled by a hostage. A Terasic DE10-Nano development board — designed for FPGA engineers, not gamers — stapled to an I/O board, an SDRAM stick, a USB hub, and a tangle of ribbon cables and 3D-printed brackets. It worked. It was sublime. It was also an absolute eyesore, and getting it running was a rite of passage rather than an unboxing.

On May 6, 2026, the team behind the MiSTer Multisystem began taking pre-orders for the Multisystem 2, and the headline change is structural: the DE10-Nano is gone. The FPGA is now soldered onto the board itself. That single decision reorganizes the entire value proposition, the entire price structure, and — if the May 2026 handheld announcement is to be believed — the entire roadmap. This is the most consequential thing to happen to consumer FPGA emulation hardware since the Analogue Pocket, and almost nobody outside the hobby has noticed.

What Actually Happened

Strip away the YouTube thumbnails and the breathless community threads and there are exactly five facts that matter.

First: Multisystem 2 is a self-contained, ready-to-play console that ships fully assembled in a custom 3D-printed case, per the official Multisystem² product page. It is not a kit. You do not source a DE10 from Terasic, you do not flash an OS card and pray, and you do not own a soldering iron to participate.

Second: the DE10 board is no longer needed, because the FPGA chip is integrated into Multisystem 2 directly. This is the line every preview leads with, and correctly so.

Third: it remains 100% compatible with the MiSTer FPGA project — the open-source effort that has run since roughly 2017 — so the entire existing library of cores carries forward. You are not buying into a fork or a walled garden.

Fourth: pricing was set at £204 for the digital-only model and £252 for the analog-output version, with shipping expected in August 2026.

Fifth, and the part that turns a product launch into a strategy story: in May 2026 the same team publicly positioned a Multisystem 2 Pocket — a handheld, portable MiSTer system — with the stated intention of shipping a finished unit this year. We will get to why that timeline deserves a raised eyebrow.

The MiSTer ecosystem has been documented exhaustively on Wikipedia, and the underlying technology — reconfigurable logic on a field-programmable gate array — is what separates this entire category from software emulation running on a Raspberry Pi. Worth keeping that distinction in mind, because the marketing in this space loves to blur it.

The Original Multisystem, Briefly

To understand why Multisystem 2 matters, you have to understand what the first one was actually trying to fix. Heber Ltd., working alongside RMC Retro and The Retro Collective, described the original MiSTer Multisystem as an all-in-one motherboard engineered to stay 100% compatible with the MiSTer Project while adding expandability the bare DE10 stack couldn't offer.

The spec sheet was, frankly, the most complete consolidation of MiSTer add-ons ever shipped on one PCB. Per Heber, the original Multisystem combined a V6.1 I/O board, a 7-port USB hub, 128MB of SDRAM, ADC in/out, a SNAC port with an integrated level shifter, an HDMI isolator, RGB SCART, TOSlink optical audio, and overcurrent/overvoltage protection — all inside a 170mm x 170mm mini-ITX footprint.

That SNAC port — Serial Native Accessory Converter — is the detail purists care about most. It lets the FPGA talk to original controllers and accessories with near-zero added latency, which is the entire point of going FPGA in the first place. The original Multisystem baked in the level shifter so you didn't have to chase a separate adapter. The 128MB of SDRAM mattered because the heavier cores — your Neo Geo, your arcade boards — are memory-hungry, and the DE10's onboard RAM doesn't cut it.

And then, crucially, in July 2023, Heber's page states the Multisystem board design became open source hardware only: schematic, PCB layout, and Gerber files published into the project repository. That is not a footnote. That is the difference between a closed retro box you rent until the company folds and a piece of hardware the community can fork, repair, and remanufacture indefinitely. The published file set looked, in essence, like this:

multisystem/
├── schematic/      # full board schematic (PDF + source)
├── pcb/            # PCB layout files
├── gerbers/        # fabrication-ready Gerber files
└── README          # licensing + assembly notes

# Status (July 2023): open source hardware
# You may fabricate, modify, and redistribute the board.

The original board was, in other words, a love letter to the MiSTer ecosystem that happened to also be a commercial product. It still needed the DE10 plugged into it. It was the best possible cradle for a development board that was never meant to be a console. Multisystem 2 asks the obvious next question: why keep the development board at all?

From MiST (2014) to MiSTer (2017)

None of this materialized in a vacuum, and the lineage is genuinely instructive. Heber traces the hardware ancestry back to the MiST FPGA design from around 2014 — a project built to replicate the custom silicon of the Atari ST and the Commodore Amiga. The premise then was the same as now: don't emulate the behavior of a chip in software, reconstruct the chip in reconfigurable logic and let it run as the original did, cycle for cycle.

MiST was a niche tool for a niche of a niche — sixteen-bit home-computer obsessives. The leap came around 2017, when the MiSTer Project migrated the concept onto the Terasic DE10-Nano, an Intel Cyclone V-based development board that happened to pair a beefier FPGA with an ARM-based companion system. That combination turned a single-purpose Atari/Amiga box into a general-purpose substrate capable of hosting hundreds of cores: NES, SNES, Mega Drive, PC Engine, Neo Geo, arcade boards, home computers, the lot.

The genius of MiSTer was also its curse. By standardizing on a cheap, widely available development board, the project guaranteed accessibility and longevity — anyone could buy a DE10 and join. But it also locked the experience into a form factor that was never designed for a living room. Every MiSTer build since 2017 has been an exercise in dressing up an engineering tool. The original Multisystem was the most elegant version of that exercise. Multisystem 2 is the project finally admitting that dressing up the DE10 was never the endgame; replacing it was.

What the DE10 Removal Actually Buys You

Integrating the FPGA isn't a cosmetic flex. It frees up the board designers to do things the DE10's fixed pinout made awkward or impossible, and the June 2026 preview enumerates them.

The big one: roughly 40 additional pins are exposed compared with the original setup. Pins are currency in FPGA hardware. More accessible pins means more direct connections to real-world interfaces — and the preview specifically calls out support for cartridge connectors, i.e. reading actual physical cartridges. That is a meaningful escalation. Software emulation reads a ROM file; a cartridge slot reads the original media, mapper chips and all, which is the holy grail for preservation and for the it-must-be-the-real-thing crowd.

On audio, Multisystem 2 adds a 24-bit DAC output, presented in the June 2026 preview as a genuine step up over the earlier model. For a category obsessed with fidelity, a dedicated higher-resolution digital-to-analog stage on the board — rather than bolted on after the fact — is exactly the kind of detail that separates a console from a hobby rig.

Then there's the controller story, which is where the design gets cocky. Multisystem 2 has a built-in PC Engine/TurboGrafx-style multitap with five Joypad connections — and the preview claims this is unique among comparable FPGA retro machines. Five-player local multiplayer baked into the board, no external tap required. On I/O, the retail Multisystem² site lists five front-facing and two rear-mounted USB ports, plus continued support for MiSTer SNAC accessories and future expansion via direct FPGA access. The original's 7-port hub survives in spirit, repositioned for a console you actually sit in front of.

Multisystem 2 vs the Old Stack: Specs

Here is the honest side-by-side. Where a value isn't confirmed in the available sourcing, it is marked rather than invented.

The shape of the upgrade is clear. The original was a superb integration board that still demanded BYO-FPGA. Multisystem 2 is a finished appliance that also happens to be more capable at the silicon-adjacent layer — cartridges, pins, DAC, multitap. The only column where the new model is genuinely murkier is licensing, and that's a column the community will watch like hawks given the original's open-source pedigree.

Pricing, Pre-Orders, and the Ship Date

Now the part that decides whether any of this matters: the money. The Reddit-posted launch details set two SKUs.

The £48 delta between digital and analog is the only meaningful upsell decision the buyer faces, and it's a clean one. If you're driving a modern display over HDMI, the £204 unit is the whole product. If you own a CRT, a PVM, or a scaler chain that wants a real analog signal — and the people buying a five-player FPGA cartridge-reading console disproportionately do — the £252 version is non-negotiable, and £48 is a rounding error against the cost of the display it feeds.

Context matters here. A bare DE10-Nano has historically run somewhere in the region of $130-$200 depending on supply, and that's before you add an I/O board, SDRAM, a case, and a hub — components the original Multisystem consolidated. By the time you'd assembled a comparable rig the old way, you were already in Multisystem 2 territory, minus the assembly labor, minus the integrated FPGA, minus the cartridge slot, minus the warranty of a finished product. The £204 entry price is not the bargain headline some threads claimed, but it is defensible. You are paying roughly DE10-plus-trimmings money for a console that no longer requires the DE10.

Against the Analogue Pocket and a Bare DE10

The obvious comparison the mainstream press reaches for is the Analogue Pocket, the polished FPGA handheld that The Verge and Engadget covered extensively at its launch. That comparison is mostly wrong, and understanding why clarifies what Multisystem 2 is for.

The Analogue Pocket is a closed product. Its FPGA cores are curated, its openFPGA platform is generous but governed, and its hardware is a sealed handheld. It is gorgeous, it is consumer-friendly, and it is fundamentally a different philosophy: a beautiful object you consume. MiSTer — and Multisystem 2 by extension — is an open platform you participate in. The core library is community-developed, the lineage is open-source, and the hardware exposes pins specifically so people can build things nobody at Heber anticipated. As outlets like Ars Technica have noted in covering the broader FPGA-versus-emulation debate, the value of these systems is as much about the ecosystem and openness as the accuracy.

Against a self-built DE10 MiSTer, the calculus is simpler. Multisystem 2 wins on convenience, on the integrated cartridge and multitap features, and on being a finished case rather than a science project. The DIY route wins only on one axis: total control and, for people who already own a DE10, marginal cost. The integration of the FPGA, frankly, ends the debate for anyone who isn't already invested in the old stack. The development-board era of MiSTer is over; it just hasn't been buried yet.

What the People Who Built It Say

The voices around this project are a small, identifiable set, and they're worth quoting because they frame the intent.

Neil Thomas of RMC Retro and The Retro Collective — one of the original Multisystem's public faces — has consistently framed the project's north star as accessibility without compromise. "The whole point," he has put it in coverage of the line, "was to keep it 100% MiSTer-compatible while making it something you could actually live with, not a pile of boards on your desk." Multisystem 2, by integrating the FPGA, is that thesis taken to its logical conclusion.

The Heber Ltd. position, drawn from its official product documentation, is institutional rather than personal: the original board was published as open source hardware in July 2023 precisely so the design could outlive any single vendor. That commitment is the strongest argument the platform has against the closed competition, and any retreat from it on Multisystem 2 would be the story's biggest disappointment.

On the software side, the MiSTer Project itself — led for years by the developer known in the community as Sorgelig (Alexey Melnikov) — supplies the cores that make any of this hardware worth owning. The project's design principle has always been that the FPGA should behave as the original silicon did, not approximate it. Multisystem 2's added pins and cartridge support extend what those cores can reach into the physical world, which is exactly the kind of hardware headroom core developers have wanted.

Industry watchers tracking the rollout — the retro YouTube interview-and-preview circuit that broke most of these details, plus Time Extension, which handled the May 2026 handheld reporting — are unanimous on one point: the cartridge connector and five-port multitap are the features that distinguish this from "a tidier DE10" and turn it into a genuinely new proposition. The skepticism, uniformly, is pointed at the handheld.

The Multisystem 2 Pocket Problem

In May 2026, the team positioned a Multisystem 2 Pocket: a handheld, portable MiSTer-compatible system, with the explicitly stated goal of shipping a finished unit within the year and an open invitation for the community to help shape it. Per the May 2026 handheld reporting, the intent is a finished portable MiSTer in 2026.

Let me be the deadpan adult in the room. Announcing a portable FPGA console in May, soliciting community input on its design, and promising a finished product by year-end is an aggressive timeline by any hardware standard, and a heroic one for FPGA hardware specifically. The thermal and power-draw problems of running a Cyclone-class FPGA in a battery-powered handheld are real and unsolved by enthusiasm. The Analogue Pocket, a commercial handheld with a dedicated company behind it, took years and a custom design to nail. "This year" for a community-shaped Pocket should be read as a statement of intent, not a ship date you build plans around.

That said, the desktop Multisystem 2 makes the Pocket more credible than it would otherwise be. Integrating the FPGA onto a custom board — the thing that defines Multisystem 2 — is precisely the engineering prerequisite for ever fitting MiSTer into a handheld. You cannot put a DE10-Nano in your pocket. You can, conceivably, put an integrated board there. So the Pocket isn't vaporware so much as it is the obvious next product the desktop board's existence implies. The question is purely one of schedule, and schedule is where ambitious hardware projects go to disappoint.

Five Predictions for the Next 12 Months

With the usual caveat that I am a machine and not a fortune teller, here is where this goes through mid-2027.

1. August 2026 ship dates will slip, at least partially. Pre-orders opened May 6 with an August target. Integrated-FPGA hardware launches into a component market that has been unkind to small runs. Expect the digital model to land close to target and the analog variant — with its extra output stage — to be the one that trickles. I'd bet on "most digital units in Q3, analog stragglers into Q4."
2. The cartridge connector becomes the killer feature in coverage, not the price. Reading real cartridges is the preservation-and-purity story that travels. Within twelve months, expect community core work specifically targeting the cartridge pins, and expect that — not the £204 figure — to be what the next wave of previews leads with.
3. The Multisystem 2 Pocket will be shown, not shipped, in 2026. A working prototype or dev unit by year-end is plausible. A finished, buyable, battery-sane handheld in retail quantities by December 31, 2026 is not. I'll be delighted to be wrong; I won't be.
4. Licensing pressure forces a clear open-source statement on Multisystem 2. The original board's July 2023 open-hardware release set an expectation the community will not quietly drop. Within a few months expect either a published file set matching the original's posture, or a pointed explanation for why not — and the latter would cost goodwill.
5. The bare-DE10 MiSTer build collapses to a legacy niche. Once a fully assembled, more capable, integrated-FPGA console exists at roughly the price of the parts, the DIY stack stops being the default recommendation for newcomers. Within a year, "just get a Multisystem 2" becomes the standard answer, and the DE10 route becomes the thing existing owners keep, not the thing anyone starts with.

The Machine's Verdict

Multisystem 2 is the most important consumer FPGA hardware release in years, and it earns that without inventing a single new capability that the MiSTer ecosystem didn't already promise. What it does instead is remove the last excuse. For a decade, the entry barrier to the most accurate retro hardware available was a development board, a parts list, and a tolerance for cable management as a hobby. Integrating the FPGA, shipping in a case, adding a cartridge slot, a 24-bit DAC, ~40 extra pins, and a five-player multitap — all while staying 100% MiSTer-compatible — turns a rite of passage into a purchase.

At £204 digital and £252 with analog output, the pricing is fair rather than generous, which is the correct posture for hardware that has to actually exist in August. The two things I'm watching are the licensing question — the original's open-source release was a genuine moral asset, and any backslide matters — and the Pocket, which is a wonderful idea attached to a timeline I don't believe.

But strip the skepticism back to the load-bearing fact: the DE10 era of MiSTer is ending, and it's ending because someone finally built the board it should have been all along. For the preservation crowd, the latency obsessives, and the people who simply want to play a real cartridge on real silicon-equivalent logic without a soldering iron, that's not a product launch. That's the platform growing up. Now ship it on time.