Type “switch 2 emulator” into Google right now and you will find forum threads, YouTube videos, and GitHub stars climbing by the day. What you will not find, as of October 10, 2026, is a single emulator that boots a commercial Nintendo Switch 2 game. Three public projects are chasing that goal anyway: Pound, oboromi, and Voland. Each one is a genuine, open-source attempt to reverse-engineer Nintendo’s newest console. None of them can play Mario Kart World yet.
That gap between search demand and actual capability is the story. “Switch 2 emulator” pulls roughly 12,100 monthly searches in the United States alone, according to keyword-volume data gathered for this piece, and the term carries low competition because almost nobody has written an honest, technical comparison of what these three projects actually do. This article fills that gap. It walks through each project’s codebase, license, platform support, and GitHub activity, lines up the Switch 2’s custom silicon against what each emulator claims to emulate, and uses the documented history of the original Switch’s emulation scene, Yuzu, Ryujinx, and the Nintendo lawsuits that followed, to put a realistic timeline on when any of this might matter for actual players.
Why “Best Switch 2 Emulator” Is Still the Wrong Question
Every other emulator comparison on the internet asks which tool runs games faster or with fewer glitches. That framing does not work for Switch 2 emulation in October 2026 because none of the three active projects, Pound, oboromi, or Voland, have published evidence of a commercial Switch 2 title booting to a title screen, let alone running in real time. The honest comparison is not about frame rates. It is about which project has built the most convincing technical foundation, which one is closest to a first boot, and which one is likely to still exist in a year.
That distinction matters for search intent. People typing “switch 2 emulator” into Google are mostly asking one of three questions: does one exist, is it safe to download something claiming to be one, and how long until a real one appears. Scam sites and fake download pages have already tried to capitalize on the search volume, bundling malware with files that claim to be a working Switch 2 emulator. The three projects covered here are legitimate open-source repositories with public commit histories, but legitimate does not mean usable. Readers searching for a way to play Switch 2 games on PC today will not find one in Pound, oboromi, or Voland, and this comparison says so plainly rather than padding out a “best of” list with projects that cannot do the thing the list implies they can do.
Meet the Three Projects Chasing Switch 2 Emulation
All three projects emerged within days of each other in June 2025, roughly in parallel with the Switch 2’s June 5, 2025 launch. That timing is not a coincidence: Nintendo’s new console ships a custom Nvidia Tegra T239 chip, codenamed “Drake,” and reverse-engineering teams had been quietly studying leaked documentation and early hardware teardowns for months before launch day. Once retail units hit shelves, three separate groups of developers went public with repositories aimed at the same target from three different angles.
Pound, hosted under the pound-emu GitHub organization, bills itself as an open-source emulator for “the Nintendo Switch 1 and 2,” written primarily in C under a GPL-3.0 license. It is the most direct spiritual successor to the Switch 1 scene: Pound’s repository explicitly targets both generations of hardware rather than starting from scratch on Switch 2 alone. oboromi, maintained by a pseudonymous developer using the handle 0xNikilite, is a Rust-based proof-of-concept that focuses narrowly on Switch 2’s low-level architecture: its own documentation describes emulating an eight-core ARMv8 CPU, 12GB of shared memory, and a stubbed-out GPU targeting Nvidia’s SM86 (Ampere) architecture. Voland takes the most unusual approach of the three, built on a custom ARM64 JIT recompiler called “Ballistic” that targets WebAssembly and x86-64, meaning its long-term ambition includes running inside a web browser. Voland’s own project notes state plainly that Switch 2 support is “planned only once the console is moddable,” meaning its current work is aimed at the original Switch, with Switch 2 as a stated future goal rather than a current capability.
Pound: The Switch 1-and-2 Hybrid Attempt
Pound is the most actively watched of the three projects by raw GitHub metrics, sitting at roughly 1,100 stars and 56 to 57 forks as of early October 2026. The repository was created on June 6, 2025, one day after the Switch 2’s retail launch, and its most recent recorded development activity lands on September 11, 2026, more than 15 months later. That is a long runway for a project that still labels itself “highly work in progress” in its own README.
What makes Pound distinct from oboromi and Voland is its dual-console ambition. Rather than treating Switch 2 as an entirely separate emulation project, Pound’s architecture is built to eventually handle both the original Switch and its successor, sharing code where Nintendo’s hardware choices overlap. The project is licensed under GPL-3.0, written in C, and documentation from a community mirror of the repository notes that Pound “actively supports Linux and Windows with the Clang compiler only,” narrowing its near-term platform ambitions to two desktop operating systems rather than the five-platform spread (Windows, Linux, macOS, Android, and browser) that mature emulators like RetroArch eventually reach.
The project’s own wiki does not shy away from managing expectations. One widely circulated warning embedded in Pound’s documentation states, in capital letters: “IMPORTANT: THIS PROJECT WILL NOT BE READY FOR A DECADE MINIMUM!” according to the project’s mirrored repository, a level of candor rarely seen in emulator README files and a useful signal for anyone hoping to play Switch 2 games on PC in the near term.
Oboromi: Eden’s Orbit and the Proof-of-Concept Approach
Oboromi has edged ahead of Pound on raw star count, with roughly 1,300 stars and 55 forks as of early October 2026, despite being created three days later, on June 9, 2025. The project’s creator, 0xNikilite, describes it explicitly as “a proof-of-concept project for a possible Nintendo Switch 2 emulator” rather than a committed, production-bound emulator, a framing that has stuck through more than a year of development. Coverage of the project from The Gamer’s reporting on Switch 2 piracy and emulation status describes oboromi in similar terms, as a “work-in-progress emulator foundation for the Nintendo Switch 2” that has “not demonstrated running Switch 2 ROMs.”
Technically, oboromi is narrower and more honest about its scope than Pound. Its documentation describes a working scaffold that emulates an eight-core ARMv8 CPU and 12GB of shared memory, matching the Switch 2’s real-world Cortex-A78C core count and LPDDR5X memory pool, alongside a stubbed GPU implementation targeting Nvidia’s SM86 (Ampere) instruction set. “Stubbed” is the operative word: the GPU layer exists to let other subsystems boot far enough to be tested, not to render actual Switch 2 graphics. A separate report from iXBT Games, citing the project description, put it plainly: “the foundation of the Switch 2 emulator is being developed,” not a finished product.
Oboromi’s development is also mirrored on git.eden-emu.dev, the self-hosted Git infrastructure run by the team behind Eden, the community fork that has become one of the leading original-Switch emulators since Nintendo’s 2024 legal campaign scattered the scene. That mirror relationship does not mean Eden’s team wrote oboromi, but it does suggest informal overlap between the developers keeping the Switch 1 emulation scene alive and the earliest people probing Switch 2’s internals. The most recent oboromi commit referenced in public records lands on September 7, 2026, with the Eden-hosted mirror syncing as recently as October 9, 2026, the day before this article was researched.
Voland: The Browser-First Long Shot
Voland is the least mature of the three by any public measure, and it is also the only one explicitly not targeting Switch 2 yet. Its GitHub organization describes Voland as “an open-source Nintendo Switch emulator that runs in the browser,” built on a custom ARM64 JIT recompiler the team calls Ballistic, designed to compile to WebAssembly as well as native x86-64 code. That browser-first architecture is unusual in emulation circles, where most serious projects target native desktop performance first and treat browser ports, if they happen at all, as a much later convenience layer.
Voland’s own project notes state that “Switch 2 support is planned only once the console is moddable,” a direct acknowledgment that the team does not yet have the firmware access or hardware documentation needed to start serious Switch 2 work. The project page is dated September 24, 2026, with activity recorded as recently as October 9, 2026, meaning Voland is both the newest publicly documented entrant among the three and the most upfront about not actually being a Switch 2 project yet in any functional sense. Reliable, independently verified GitHub star and fork counts for Voland were not available at the time of research, a gap this article flags rather than papering over with an invented number.
Why the Tegra T239 Is a Much Harder Target Than the Original Switch
Understanding why none of these three projects have booted a game requires understanding what they are actually up against. The Switch 2 runs on a custom Nvidia Tegra T239 system-on-chip, codenamed “Drake,” pairing eight ARM Cortex-A78C CPU cores with an Nvidia Ampere-architecture GPU built from 12 streaming multiprocessors and 1,536 CUDA cores. Nintendo quotes roughly 3.07 teraflops of graphics performance docked and about 1.71 teraflops in handheld mode. The system ships with 12GB of LPDDR5X memory over a 128-bit interface, delivering around 102GB/s of bandwidth in docked mode, backed by 256GB of UFS 3.1 internal storage and a 7.9-inch, 1920×1080 display capable of variable refresh rates up to 120Hz.
Every one of those specs represents a reverse-engineering problem. The original Switch’s Tegra X1 was already a known quantity by the time Nintendo shipped it in March 2017, since the same chip had already appeared in the Nvidia Shield set-top box and Android tablets, giving early emulator developers a head start on understanding its instruction set and driver behavior. The Tegra T239 has no such public precedent. Its Ampere-based GPU architecture is closer to Nvidia’s desktop RTX 30-series lineage than to anything in a prior game console, and Nintendo has not released developer documentation to the public. Every detail that oboromi’s stubbed GPU layer or Pound’s CPU core model gets right has to be inferred from firmware dumps, driver behavior, and painstaking trial and error, not read out of a datasheet.
What a Real Switch 2 Emulator Would Still Need to Solve
Even if Pound, oboromi, or Voland eventually nails the raw CPU and GPU instruction modeling, booting a Switch 2 title to a title screen is only the first of several hard problems, not the last. The console relies on Nintendo’s own cryptographic key hierarchy to decrypt game cartridges and system firmware, the same general approach the original Switch used, which means legitimate emulation still requires a user to supply their own legally dumped keys and firmware from hardware they own. Nothing about Pound’s or oboromi’s current codebase changes that requirement, and no credible project in this space has claimed otherwise.
Beyond decryption, the Switch 2 leans harder on Nvidia’s DLSS-style upscaling and frame generation than any previous Nintendo console, using dedicated Ampere tensor cores to reconstruct a sharper image from a lower native render resolution. Faithfully reproducing that behavior on a PC’s GPU, whether Nvidia, AMD, or Intel, is a materially different problem than the straightforward shader translation that let PS2 and GameCube emulators reach high compatibility years ago. Oboromi’s own documentation stops at a “stubbed” GPU target precisely because this layer, mapping Ampere-specific tensor and shader behavior onto whatever graphics card the end user owns, is still unsolved in public code. Pound’s team has not published equivalent GPU documentation at all, and Voland’s current scope does not yet touch Switch 2 hardware, so none of the three projects has a public answer for DLSS emulation today.
There is also the matter of online services and anti-piracy telemetry. Nintendo Switch Online integration, software updates distributed over the internet, and any server-side validation Nintendo has built into Switch 2 titles are all invisible to outside developers until someone documents them through packet analysis or firmware teardown, work that takes months even for well-resourced reverse-engineering teams and carries its own legal exposure under the same DMCA framework that ended Yuzu. Every one of these layers, decryption keys, DLSS reconstruction, and online validation, sits on top of the basic CPU and GPU emulation that Pound and oboromi are still working through, which is part of why Pound’s own README settles on a timeline measured in decades rather than months.
Specs and Status Comparison: Pound vs. Oboromi vs. Voland
The table below lines up every verifiable technical and project-management data point across the three repositories, gathered directly from GitHub as of early October 2026.
| Metric | Pound | Oboromi | Voland |
|---|---|---|---|
| GitHub organization / repo | pound-emu/pound | 0xNikilite/oboromi | voland-emu |
| GitHub stars (as of Oct. 2026) | ~1.1K | ~1.3K | Not reliably verified publicly |
| Forks | 56–57 | 55 | Not reliably verified publicly |
| Primary language | C | Rust | Not publicly specified in detail |
| License | GPL-3.0 | GPL-3.0 | Not publicly confirmed |
| Repository created | June 6, 2025 | June 9, 2025 | Active as of Sept. 24, 2026 (creation date unconfirmed) |
| Most recent activity | Sept. 11, 2026 | Sept. 7, 2026 (mirror synced Oct. 9, 2026) | Oct. 9, 2026 |
| Target console(s) | Switch 1 and Switch 2 | Switch 2 only | Switch 1 now; Switch 2 “planned” |
| Stated platform support | Windows and Linux (Clang) | Not fully documented publicly | WebAssembly (browser) and x86-64 native |
| CPU emulation claimed | Not fully documented publicly | 8-core ARMv8, matching Switch 2’s Cortex-A78C count | Custom ARM64 JIT recompiler (“Ballistic”) |
| GPU emulation claimed | Not fully documented publicly | Stubbed Nvidia SM86 (Ampere) target | Not applicable to current Switch 2 scope |
| Memory model claimed | Not fully documented publicly | 12GB shared memory model | Not applicable to current Switch 2 scope |
| Commercial games booting | None documented | None documented; explicitly “cannot run commercial software” | None for Switch 2 (not yet in scope) |
| Project’s own risk disclosure | “Will not be ready for a decade minimum” | Described by creator as a “foundation” being developed | Switch 2 work contingent on console being “moddable” |
Benchmarks: What “Progress” Actually Looks Like Right Now
Traditional emulator benchmarks measure frames per second against specific games. That metric does not exist yet for Pound, oboromi, or Voland, because no commercial Switch 2 title has been reported booting on any of the three. The meaningful benchmark in October 2026 is architectural: how much of the real Tegra T239 has each project modeled, and how far does execution get before it stalls.
Cross-referencing GitHub’s own repository data, the community tracking site GitTrend, and independent reporting from outlets covering the emulation scene gives a consistent picture across three separate sources. GitHub’s own organization pages confirm Pound’s roughly 1,100 stars and oboromi’s roughly 1,300 stars as of early October 2026. GitTrend’s independently scraped star-history snapshots, taken slightly earlier in September 2026, showed both projects a few dozen stars lower, consistent with steady organic growth over several weeks rather than a viral spike. Reporting from outlets covering the emulation community, including coverage that quoted the Pound project’s own “not ready for a decade” disclaimer, corroborates that neither project has claimed, let alone demonstrated, a commercial game boot. Where those three source types agree is more useful here than a frame-rate chart would be: this is pre-alpha reverse-engineering work, not a feature race.
Context from the original Switch’s emulation history makes the timeline gap concrete. The first Nintendo Switch launched on March 3, 2017. Ryujinx’s development began in December 2017, and by early February 2018, roughly eleven months after launch, it was already showing a boot logo for the Switch port of Puyo Puyo Tetris, with several more commercial titles booting by April and May 2018. Yuzu followed a similar curve, confirming its first commercial game boots, including 1-2-Switch, in late April 2018, about 14 months after the console’s launch. The Switch 2 launched on June 5, 2025. As of this article’s October 10, 2026 research date, that is more than 16 months of real-world time, longer than it took either Ryujinx or Yuzu to reach first boot on the original hardware, and none of the three known Switch 2 projects have reported an equivalent milestone.
Lessons From the Switch 1 Emulation Wars
The slower pace is not purely a technical story. It is also a legal one, and the fate of the original Switch’s two flagship emulators looms over everything happening with Pound, oboromi, and Voland today. Nintendo sued Tropic Haze LLC, the company behind Yuzu, in February 2024, alleging the emulator violated the DMCA’s anti-circumvention provisions by helping users bypass Switch encryption. Tropic Haze settled on March 4, 2024, agreeing to pay Nintendo $2.4 million in damages, accept a permanent injunction barring any future work on Switch-circumvention software, and pull Yuzu’s source code from GitHub entirely.
Ryujinx did not survive much longer. On September 30, 2024, Nintendo contacted the project’s lead developer, known publicly as gdkchan, reportedly offering an agreement that required him to stop working on the emulator and remove the Ryujinx GitHub organization and related assets. The organization’s repositories were taken down by October 1, 2024. Community successors, including Eden, Citron, and a Ryujinx-derived fork now maintained as Ryubing, kept Switch 1 emulation alive, but several of those projects have since relocated away from GitHub to self-hosted infrastructure after further takedown activity swept through the scene.
That history is directly relevant to Pound, oboromi, and Voland. All three projects currently live openly on GitHub, something Yuzu and Ryujinx could no longer do by late 2024. The most plausible explanation is not that Nintendo has overlooked them, but that none of the three has yet built something capable of running a pirated commercial game, which was the specific legal trigger Nintendo cited against Yuzu. If any of these projects crosses that line, the same DMCA playbook that erased Yuzu and Ryujinx from GitHub in 2024 is still available to Nintendo’s legal team, and nothing about the current environment suggests the company has softened its stance.
Pricing: What It Actually Costs to Chase Switch 2 Emulation Today
All three emulator projects are free, open-source software distributed under permissive or copyleft licenses, so there is no purchase price to compare in the traditional sense. The more useful pricing comparison is the cost of the hardware people are weighing against each other: buy the real console, or gamble time on software that cannot yet play anything.
| Option | Price | What you actually get today |
|---|---|---|
| Nintendo Switch 2 console (official) | $449.99 at launch; raised to $499.99 as of September 1, 2026 | Full access to the entire Switch 2 game library, day one |
| Pound (emulator) | Free, GPL-3.0 | A work-in-progress codebase; no commercial games boot |
| Oboromi (emulator) | Free, GPL-3.0 | A CPU/memory/GPU-stub proof of concept; no commercial games boot |
| Voland (emulator) | Free, license unconfirmed | A Switch 1-focused JIT recompiler; Switch 2 work not yet started |
| Gaming PC capable of someday running a mature Switch 2 emulator | Roughly $1,200–$2,500 for a modern Nvidia RTX-class rig | Hardware headroom for whenever (if ever) a functional build ships |
For nearly everyone reading this in October 2026, the console remains the only path to actually playing Switch 2 games, even with the price increase to $499.99. Nintendo has sold 19.86 million Switch 2 units worldwide as of March 31, 2026, according to the company’s own financial disclosures, less than a year after the console’s debut, and that installed base keeps growing regardless of what happens in these three GitHub repositories.
Who Should Actually Pay Attention to These Projects Right Now
Because none of these three emulators can play games, “best use case” has to be reframed around who genuinely benefits from following or contributing to them today.
- Homebrew and reverse-engineering enthusiasts who want to understand the Tegra T239’s instruction set and memory architecture will find oboromi’s documented CPU and memory scaffolding a useful, if incomplete, reference.
- C and systems programmers looking for an open-source project to contribute low-level code to, rather than UI or game-specific patches, are a better fit for Pound’s architecture than for most mature, feature-complete emulators that mainly need compatibility patches now.
- Switch 1 emulator users on Eden, Citron, or Ryubing who want to track how their existing scene connects to Switch 2 work should watch oboromi specifically, given its mirror on Eden’s self-hosted infrastructure.
- Web developers interested in WebAssembly-based emulation have essentially one project to watch in this space: Voland’s Ballistic recompiler is a genuinely unusual architecture choice worth studying even before it supports Switch 2.
- Journalists, analysts, and competitive intelligence teams tracking how quickly console security holds up post-launch have a clean comparison point: the original Switch took roughly 11 to 14 months to reach first commercial boot, and Switch 2 has now passed 16 months with zero equivalent milestones reported.
- Casual players who just want to play Switch 2 games are not well served by any of these three projects today. Buying or borrowing the actual console remains the only realistic option in October 2026.
A Safe On-Ramp: How to Actually Explore These Projects Without Getting Burned
Given how early this scene is, the biggest practical risk is not legal exposure, it is malware. Search traffic for “switch 2 emulator” has already attracted fake download pages bundling malicious installers with files claiming to be a working emulator. Anyone curious enough to explore Pound, oboromi, or Voland firsthand should treat this less like installing a finished app and more like checking out a research codebase, because that is what it actually is.
- Go directly to the official GitHub organization pages, pound-emu/pound, 0xNikilite/oboromi, or voland-emu, rather than following links from YouTube descriptions or forum posts.
- Read the README and wiki pages in full before doing anything else; both Pound and oboromi disclose their limitations clearly, and that disclosure is the single most reliable signal of a legitimate project in this space.
- Check the commit history and issue tracker for real, recent activity rather than relying on star count alone, since stars can be inflated and do not reflect functional progress.
- If you plan to build from source, do it in an isolated environment (a VM or a spare machine) rather than your primary PC, standard practice for any early-stage, unaudited native code.
- Do not download prebuilt binaries claiming to run commercial Switch 2 games from any source; no legitimate project has published one, and any file claiming otherwise should be treated as a red flag.
- If you are coming from the Switch 1 scene (Eden, Citron, Ryubing, or older Ryujinx forks), understand that none of your existing configuration, firmware dumps, or save files will transfer to Switch 2 emulation, because the underlying CPU architecture, GPU, and memory layout are fundamentally different chips.
- Follow the projects’ GitHub Discussions or linked community channels for status updates rather than third-party aggregator sites, which have a track record of overstating progress for clicks.
- Reassess every few months rather than checking daily; based on the original Switch’s emulation timeline, meaningful progress on a project this early is measured in months and years, not days.
Pros and Cons of Each Project
Pound
- Pros: Dual-targets Switch 1 and Switch 2 in one codebase; transparent about its decade-long timeline; active commit history through September 2026; permissive GPL-3.0 licensing encourages contributions.
- Cons: Narrower documented platform support (Windows and Linux only, via Clang); less publicly detailed CPU/GPU architecture documentation than oboromi; no commercial game compatibility of any kind.
Oboromi
- Pros: Highest GitHub star count of the three (~1.3K); most detailed public documentation of Switch 2’s CPU core count and memory model; informal ties to Eden’s established Switch 1 infrastructure; written in memory-safe Rust.
- Cons: Explicitly a proof-of-concept, not a production roadmap; GPU layer is a stub, not a renderer; creator’s own framing suggests a “foundation” stage with no committed completion timeline.
Voland
- Pros: Architecturally the most novel of the three, with a browser-capable WebAssembly target; newest and most actively updated as of early October 2026; clear, honest public statement that Switch 2 work has not started.
- Cons: Not currently a Switch 2 project at all, only a stated future intention; least publicly verifiable metrics (stars, forks, license) of the three; Switch 2 support is explicitly conditional on the console becoming “moddable,” a milestone with no announced date.
Migrating Expectations: What Switch 1 Emulator Users Need to Unlearn
Players coming from a mature Switch 1 setup on Eden, Ryubing, or an older Ryujinx build need to recalibrate what “emulator” means in this context. On the original Switch, the hard reverse-engineering work was finished years ago; today’s debates are about compatibility patches, shader caching, and performance tuning on known hardware. None of that applies yet to Switch 2. There is no BIOS or firmware dump to legally source and drop into a configuration folder, no save-file compatibility to worry about, and no shader cache to transfer, because there is no functional renderer for any of these three projects to cache shaders for in the first place.
The most useful mental shift is treating Pound, oboromi, and Voland the way the emulation community treated Ryujinx and Yuzu back in late 2017, not as finished tools, but as early reverse-engineering efforts whose public code happens to be visible while the real work, understanding an undocumented Ampere-based GPU and an unfamiliar memory architecture, is still being done. The original Switch needed roughly a year of that kind of work before Puyo Puyo Tetris booted on Ryujinx. The Switch 2’s T239 chip is a more complex, less publicly understood target, and 16-plus months in, the scene has not yet cleared that same bar.
The Verdict: Three Legitimate Projects, Zero Playable Games
Judged purely on data rather than hype, oboromi currently has the strongest technical documentation of the three, with a clearly described CPU and memory model that maps onto the Switch 2’s real specifications, plus the highest star count and an informal connection to Eden’s established Switch 1 infrastructure. Pound has the more ambitious long-term architecture, targeting both console generations in one codebase, but pairs that ambition with the single most honest disclaimer in the entire emulation scene: it will not be ready for a decade at minimum, by the project’s own account. Voland is the one to watch for a different reason entirely, its browser-capable WebAssembly approach is architecturally unlike anything else in console emulation, but by its own team’s admission, it is not yet a Switch 2 project in any functional sense.
None of that changes the bottom line for anyone searching “switch 2 emulator” hoping to play games today. The original Nintendo Switch took roughly 11 to 14 months after its March 2017 launch for Ryujinx and Yuzu to boot their first commercial titles. The Switch 2 launched June 5, 2025, and more than 16 months later, across three publicly tracked projects with a combined GitHub following well into the thousands, zero commercial games have been reported booting on any of them. Anyone weighing a $499.99 console against free software that cannot yet play anything is not actually choosing between two ways to play Switch 2 games. They are choosing between playing them now, or waiting an unknown number of years for software that its own developers admit is not close.
Frequently Asked Questions
Is there a working Nintendo Switch 2 emulator in 2026?
No. As of October 10, 2026, none of the three publicly known Switch 2 emulator projects, Pound, oboromi, or Voland, have demonstrated a commercial Switch 2 game booting. All three are early-stage, open-source reverse-engineering efforts rather than finished, playable emulators.
What is the difference between Pound, oboromi, and Voland?
Pound targets both the original Switch and Switch 2 in a single C-based codebase and supports Windows and Linux. Oboromi is a Rust-based proof-of-concept focused specifically on Switch 2’s CPU, memory, and a stubbed GPU layer. Voland is a browser-capable WebAssembly emulator currently focused on the original Switch, with Switch 2 support described as a future goal contingent on the console becoming moddable.
Why is Switch 2 emulation taking longer than the original Switch?
The Switch 2’s Tegra T239 chip uses an Nvidia Ampere-based GPU architecture that had no prior public release in any other device, unlike the original Switch’s Tegra X1, which developers had already studied in the Nvidia Shield set-top box before Nintendo’s console launched. That lack of public precedent makes reverse-engineering the Switch 2’s GPU and memory behavior significantly harder.
Is downloading a “Switch 2 emulator” from a random website safe?
No legitimate, functional Switch 2 emulator capable of playing commercial games currently exists, so any site advertising a working download is almost certainly distributing malware or a scam. Only the official GitHub repositories for Pound, oboromi, and Voland represent verified, legitimate development activity, and even those do not offer a playable product.
Did Nintendo shut down Yuzu and Ryujinx, and could the same happen to these projects?
Yes. Nintendo sued Tropic Haze, the company behind Yuzu, in February 2024 over DMCA anti-circumvention violations, reaching a $2.4 million settlement on March 4, 2024, that shut the project down. Ryujinx’s lead developer was contacted by Nintendo on September 30, 2024, and the project’s GitHub organization was removed by October 1, 2024. Pound, oboromi, and Voland remain publicly hosted on GitHub as of October 2026, likely because none has yet demonstrated the ability to run a commercial game, the specific trigger Nintendo cited against Yuzu.
Can I use my existing Switch 1 emulator setup (Eden, Ryubing, Citron) with Switch 2 games?
No. The Switch 2 uses an entirely different CPU core design, GPU architecture, and memory layout from the original Switch. Configuration files, firmware, shader caches, and saves from Switch 1 emulators are not compatible with Switch 2 hardware or with any of the three current Switch 2 emulator projects.
How much does the Nintendo Switch 2 cost compared to waiting for an emulator?
The Nintendo Switch 2 launched at $449.99 and rose to $499.99 as of September 1, 2026. All three emulator projects are free and open-source, but since none can currently play any games, buying the console remains the only way to actually play Switch 2 titles today.
Which Switch 2 emulator project should developers contribute to?
That depends on interest and skill set. Systems programmers comfortable in C may prefer Pound’s dual-console architecture. Rust developers interested in CPU and memory modeling may find oboromi’s narrower, more documented scope easier to contribute to. Web developers curious about WebAssembly-based emulation have essentially one option in Voland, though its Switch 2 work has not yet begun.
