[Discussion] Home Assistant Yellow Upgrades: CM5 Compatibility in HAOS 14 Boosts Performance

[Uncrowned Guard]

The Evolution of Home Assistant Yellow: Now Supports CM5 in HAOS 14

Two years ago, the computing world embraced the innovative Home Assistant Yellow. It was envisioned with a flexible design philosophy, ready to sustain and expand along its users' demands. Home Assistant wanted to create a solution where you could add storage as per your needs and easily shift firmware from Zigbee to Matter over Thread.

Embracing the Future: Integration of Compute Module 5

With the early sample provided by Raspberry Pi, they successfully integrated the state-of-the-art Compute Module 5 (CM5) compatibility to Home Assistant Yellow. The introduction of this feature into the upcoming Home Assistant OS 14 will provide current and prospective users the option to augment their device performance. However, it is worth mentioning that CM4 is still potent enough for most Home Assistant Users' needs.

Through participation in the Open Home Foundation, Home Assistant continuously strives for privacy, choice, and sustainability in the smart home industry. With Home Assistant Yellow, they have managed to fulfill these three key aspects, and the integration of CM5 reinforces the viability of broader choices and long-term sustainability.

The Advantages of Compute Modules

Designing the Home Assistant Yellow around Raspberry Pi’s Compute Module platform was a strategic move. This innovative platform gives users the option to amplify the RAM, add eMMC, built-in Bluetooth, or upgrade speed- without necessitating a complete replacement of the Home Assistant Yellow. All users are required to do is slot in a new module. This feature enhances the repairability and life span of the product.

Despite being designed around Compute Module 4 (CM4), the dream of compatibility with its future successors was always alive. This dream is now a reality as the Yellow is compatible with CM5 after comprehensive firmware updates and rigorous testing of the early hardware.

The Benefits of CM5

Experimenting with the latest hardware is always a thrill. It's especially rewarding when discovering new applications for the hardware it succeeds. The utilitarian value of old Pi products isn't diminished since these devices make excellent Wyoming satellites.

Speaking of typical usage, there isn't a marked difference between a Green, CM4, or CM5 when running automation or connecting devices for an average home. Yet, for power users, the CM5 might entirely revolutionize performance. The CM5 has been shown to improve ESPHome compilation times almost three times, translating into significant time savings in larger deployments. Local speech-to-text processing is another area where the CM5 vastly outperforms the CM4, specifically when using Assist fully locally.

Installation on CM5

Installation on the CM5 is a dubious task compared to the CM4 due to firmware differences preventing booting off USB 2.0 devices. Though complex, the rpiboot method of the Home Assistant Yellow opens the door to vast possibilities. Upgrading from an existing system to CM5 can usually be done simply by dropping in a replacement.

The Continued Relevance of CM4

Despite the new developments, the Home Assistant Green or Yellow with CM4 remains sufficiently formidable. According to our analytics, a significant number of Home Assistant users are functioning on Pi 3 hardware. Home Assistant always receives updates that enhance speed across all hardware, allowing devices to remain functional for extended periods.

In light of this, the Raspberry Pi fraternity has pledged their full support for CM4 until 2034, underscoring a commitment to continuously manufacture them.

New Features in HAOS 14

In the upcoming release of Home Assistant OS 14, several other additions are set to roll out. One such is support for the Hailo-8 AI accelerator - the AI acceleration element found in the recently launched Raspberry Pi AI Kit or Raspberry Pi AI HAT+. This advanced addition allows Pi 5 users to delegate AI processing tasks, such as object or person detection, to these efficient add-ons.

View full article