"I'm crying, finally Mountain Lion on M-series chips": one person moved everything from Tiger to Big Sur onto Apple silicon, gained 330 stars in 3 days

Virtual Intel Mac is a free Mac app that lets Apple-silicon Macs run Intel-version Mac OS X 10.4 Tiger through macOS 11 Big Sur, with graphics acceleration, so old software like iWork '09, Aperture, Final Cut Pro 7, and Front Row runs again. Repo created in Oct 19, it only reached 73 stars in the first two weeks; in Nov 3 it shipped 2.0, hit 401 three days later, and by that afternoon 464 (GitHub, 11-06). It took off because it did something nobody had managed before, came with screenshots nostalgia fans couldn't resist sharing, and carried the author's reputation from a previous project. Of those three, the first two can be learned; the third has to be built over time.

A 2:30 a.m. post

In 2026 Nov 4, early morning 2:39 (UTC), X user Diego (@1000kilobytes) posted: "OMG I'm going to cry, thanks @nfzeroxapps for making this Intel Mac VM app, I can finally run OS X Mountain Lion with graphics acceleration on Apple silicon." He attached three screenshots of Mountain Lion running on a new Mac.

This post got 51.9k views (X, 11-06). One reply under it: "Finally, I can build iOS 6 apps on my M1 without relying on that 2009-year-old Mac."

The day before, author nfzerox had released 2.0. 1.0 only supported Mojave, Catalina, and Big Sur, and sat quiet for two weeks; 2.0 added everything from Tiger to High Sierra in one go, bringing back the translucent menu bar, the reflective Dock, and the ripple effect when adding widgets to Dashboard. That's what Diego screenshotted.

What it actually is

A double-click, free, MIT-licensed Mac app. You pick an OS version, and it downloads and installs the system itself, then sets up the graphics drivers and creates the user account. OS X Daily called it "surprisingly simple."

• What it runs: 10.4 to 10.13 has OpenGL acceleration; 10.14 to 11 has Metal and OpenGL acceleration; 10.4 to 10.6 can also run PowerPC software through the system's built-in Rosetta, such as AppleWorks.
• How fast: the README says graphics performance on the M4 is roughly equivalent to a 2016-year-old 15-inch MacBook Pro, and CPU performance is roughly equivalent to a 2010-year-old 11-inch MacBook Air. Good enough for old software, not for heavy games.
• Trade-offs: installing a system takes one to two hours, and installing the drivers adds 15 to 30 minutes; you can't sign in to iCloud; and you have to bypass Gatekeeper on first launch.

The hard part is at the low level. Apple's own virtualization framework on Apple silicon only runs ARM builds of macOS 12 and later. To let old Intel systems use the host GPU, the author pulled the virtual graphics driver that only shipped with Big Sur and ported it back to Mojave and Catalina, wrote a separate OpenGL passthrough for Tiger through High Sierra, and ran an x86 helper under Rosetta 2 that shares memory with the ARM build of QEMU. The README spells out each of these, with every point linked to source code.

How people handled it before

So the README's first line says "For the first time ever," and it's not hype. "Apple silicon, Intel 10.4 through 11, with acceleration, one-click install": nobody had done that before it. The iWeb site owner wrote in the thread: "I can finally use an Apple silicon Mac without keeping those old machines running."

The demand is real: since Catalina, macOS has dropped 32-bit software, Aperture got stuck on Mojave, and old photo libraries, old websites, old games, and iOS 6 development could only be run on preserved hardware. By 2026 Intel Macs are basically gone, and Rosetta 2 is expected to be removed in macOS 28. Right now more people than ever want to save an old system as a single file.

Where the stars came from

(stars from GitHub daily data; live count in the afternoon of 11-06 was 464)

Two things stand out.

Coverage alone didn't help 1.0. LockerGnome wrote a long piece in Oct 29 about running 32 apps on Mojave and running Aperture, and over the next five days stars went from 70 to 73. Mojave is "not old enough" for most people and has nothing worth screenshotting. 2.0 Only after adding the Tiger and Snow Leopard Aqua interface did the same crowd start sharing it.

People really are downloading it. 2.0's installer has been downloaded 1,398 times, and 1.0 has been downloaded 288 times (GitHub Releases, 11-06). That's the same order of magnitude as the star count, so it's not just people starring to look.

No paid promotion, and it didn't hit the HN front page. MacRumors, 9to5Mac, and AppleInsider's main sites haven't covered it. So far it has spread only within the small circles of Mac nostalgia and virtual machines.

Why the press followed

The author, nfzerox, only joined GitHub in October 2024, and his X account has posted just 15 times, but all his projects are the same kind: rescuing old things. His previous project, Virtual Mac on iPad (running a full macOS on an iPad), has 2,026 stars, was shared by SnazzyLabs' Quinn Nelson in August, got coverage on OS X Daily, and scored 24 points on HN. His other repos, launchbad-revived (bringing back Launchpad), ipod_theme, and BackToCatalina, follow the same path.

So when 2.0 launched, OS X Daily picked it up within two days. For an open-source project, "this person's last thing was good" is enough for an editor to click.

Some people weren't buying it

Someone in the MacRumors thread said: "The repo description is long and wordy, it looks like a pile of vibe-coded stuff, and I especially love the commit messages, all 'Update repo'." Others actually installed Snow Leopard: the pointer was inaccurate, one click registered as seven, auto-configuration didn't take, and the driver install disc wouldn't read.

The author doesn't shy away from AI. The README is blunt about it: if you hit a crash, point Codex or Claude Code at the repo, have it debug and fix the problem on the spot, then open a PR. For one person doing a port this deep, maintenance depends on handing troubleshooting to the user's AI. That saves a lot of time replying to issues, but whether the people who hit problems can actually fix them on their own is still an open question.

There are two more risks: it extracts Apple system components from Big Sur for the port, and I haven't seen any takedown notice yet, but that doesn't rule one out; and once Rosetta 2 is removed, the author says he "plans to update," but whether that's achievable is unknown.

Are copycats showing up

Not yet. I found no similar new repos on GitHub, just 19 forks with ordinary contributions.

The reason is simple: the barrier isn't the idea, it's the low-level porting work. UTM has been able to run old OS X for a long time, everyone knows the missing piece was acceleration, but nobody had built it.

If it were you

C: Don't build a second one. Copy how it picks its topic and ships releases. This spot is held by a low-level breakthrough that's hard to replicate, and the author already has press relationships; my rough estimate is the final star count will land somewhere in the hundreds to two thousand, similar to his last project. If you want stars, pick a "gap left by a platform generation change" and launch it the same way.

Who it suits: people who know low-level territory like operating systems, drivers, emulators, and reverse engineering, who are nostalgic players themselves, and who can use AI coding to grind through a port that used to need a small team.

Who it doesn't suit: anyone who wants to copy it and build yet another Intel Mac VM (the spot is taken; if you want to improve on it, submitting PRs to it beats starting a new library). Also anyone who can't be "first" and can only be "a bit better" (UTM has 36k stars and still didn't catch this wave).

How to do it: find something that a platform change has cut off but that people still can't live without. The nearest right now is x28 Mac software that stops running once macOS 2 removes Rosetta 86. Looking further out, there's acceleration for Classic Mac OS and PowerPC systems, old iOS versions and old iPod games, and old drivers and games that don't run on Windows on ARM. Make something that does what nobody has done before, packaged as a double-click install, not a command-line tutorial.

How to get users: first post in the oldest, most knowledgeable forums and on X yourself (MacRumors, the relevant subreddit), and state in the first line of the README what "first" you achieved. Release big versions around "support for a few more systems and the chance to see what that era looked like," not around fixed bugs, and pair every release with a screenshot people want to share. Vertical outlets like OS X Daily will find you on their own, as long as your name has come up in the community before.

When to get in before the window closes: macOS 28 is expected to be announced in June 2027 and released in the fall. The months when Rosetta 2 actually gets removed are when "my old software won't open" will peak, so the tool should be ready before then (rough estimate).

What would show you got it wrong: ① Virtual Intel Mac is ordered taken down by Apple, showing the route of porting Apple system components doesn't work; ② after MacRumors' main site, 9to5Mac, or big nostalgia YouTubers pick it up, it rockets past 5,000 stars, showing the ceiling for this kind of project is higher than expected and it's worth betting directly on the same theme; ③ UTM adds graphics acceleration for old macOS within two or three months, showing big projects will quickly absorb this kind of gap.