Virtual machine runs Googlebook OS on Apple Silicon Macs

Skylar Taylor-Barrick has created a virtual machine tool called gbos-vm that allows the Googlebook OS to run on Mac computers with Apple Silicon chips. This tool uses the actual restore image of the Dell XPS Googlebook, a system built on Android 17 for the ARM64 architecture. Because the Dell computer uses a Snapdragon X chip that shares the same Arm architecture as Apple Silicon, the system can run directly through macOS’s hypervisor without needing heavy processor emulation, which is not possible on older Intel Macs.

Overcoming graphics and memory challenges

The main challenge in adapting the Googlebook OS to macOS was supporting hardware acceleration, which is crucial for rendering the interface smoothly and running the Chrome browser. The system relies on the Vulkan graphics API and Linux-specific memory-sharing mechanisms, both of which are not available on macOS. To address this, the project uses QEMU through the UTM components and translates Vulkan calls to Apple’s Metal API using the MoltenVK layer. The developer also adjusted shared memory management and realigned texture data on the Mac side to prevent visual errors.

AI-assisted software assembly

The project also replaces the Dell-specific vendor partition with virtual drivers from Google’s open-source Cuttlefish project. Skylar Taylor-Barrick explained on Reddit that she used several AI models, including GPT 6 Astra, DeepSeek 4.1 Flash, and Claude Opus 5.5, to help assemble and integrate the different software components of the project. This approach highlights the increasing use of artificial intelligence in complex software development tasks.

Limitations and security concerns

Despite its achievements, the virtual machine has several limitations. The display is capped at 60 images per second, Bluetooth crashes at startup unless a fake radio chip is simulated via the Android emulator, and audio performance is unreliable. Mouse cursor management is still experimental. Additionally, the system lacks a hardware security chip, forcing it to use software-based encryption. A background process related to the TPM module also crashes repeatedly, raising security concerns. The project creator advises against using it with primary Google accounts that contain sensitive data.

Installation requirements and performance

The project is available on GitHub, and installation is automated through the script ./install.sh. It requires a Mac with Apple Silicon, at least 16 GB of RAM, about 60 gigabytes of free disk space, Homebrew, Xcode command line tools, and the Android SDK and NDK. On a MacBook Pro M5, the compilation takes about five minutes, and the first startup takes approximately 45 seconds.