Sudden CPU frequency drop from 1.3GHz to 0.9GHz since v13.3.0

Homey support conceded today that there is in fact an issue with the firmware and a fix is being worked on

Good to hear that they are addressing the issue. Did they provide info on whether this is a Thread-specific issue or something broader?

They said that my Thread network is being “overloaded” and recommended that I reduce the number of devices in my network until a fix is implemented. They gave no indication on what devices were to blame other than “sockets”, despite the fact I have sockets from three different brands.

To me this is an admission that Homey Pro 2023 and Homey Pro 2026 on Firmware 13.3 is currently unable to sustain a high density Thread network.

I think that’s a pretty huge deal and I’ve asked them to clarify if this is now their official position and that 13.3 has a critical flaw.

If it’s the case, then the community needs clear guidance.

I also don’t think it’s coincidence that this has happened after major changes to how the CPU scheduler functions, despite repeated assurances on this thread that I am completely wrong.

Another suspicious drop in Ram usage this morning for no apparent reason

Last night everything was stable and working.

This happened at 8:29am. I woke up at 8:45am and most devices are offline and unresponsive.

On another thread today


Have the same problems with losing thread devices since 13.3.0 - had no problems with the last firmware before. Changed nothing in my setup

More 13.3 funny business.

85% CPU load spike.

I wasn’t even in the house, I’d gone out for a run.

Nothing on the timeline during this time.

@deejayreissue; Just like my colleagues from our Support team mentioned in their email, the issues you’re seeing on your Thread network are most likely unrelated to the update to firmware v13.3.0.

From your diagnostic report, we can see a couple of things.

First, you have a mix of IKEA, Aqara, and Eve smart plugs that together are generating more than 900 Thread messages per minute. This creates a very busy network, leaving little bandwidth for other devices to communicate.

What makes this worse is that when messages fail to reach their destination, devices automatically retry sending them. Those retries further increase network traffic, which can lead to a snowball effect where communication becomes increasingly unreliable.

We also see several errors indicating that Homey is occasionally unable to access the Zigbee/Thread radio channel. This could point to interference on the 2.4 GHz band, either within your home or from nearby wireless networks and devices.

The improvements we’re currently working on are based on these findings, but they are not related to firmware v13.3.0 itself. Specifically, we’re:

  1. Improving how the Thread network behaves when messages fail to reach their destination, helping to prevent devices from dropping off the network.

  2. Adding a device setting that allows users to adjust how frequently devices send updates, making it easier to reduce unnecessary network traffic when needed.

Regarding the memory usage pattern you observed, it appears that one of the apps running on your Homey has a memory leak. Once it reaches the maximum amount of memory it’s allowed to use, the app is automatically restarted. Without knowing which apps you have installed, it’s difficult to determine which one is responsible.

Deleted all Eve smart plugs and the Eve App. No more problems for me so far. Eve Smartplugs are always communicating because of their Energyreport I think. Nevertheless I haven’t had the thread-getting-lost-and-get-back-randomly before. Will keep watching

It’s not, false alarm. It’s something to do with Homey, definitley. Apple Home and Home Assistant doesn’t have any problems in the same network with the same devices :man_bowing:

Could you describe these problems and what is/isn’t working in Homey compared to Apple Home and Home Assistant.

Do you use any zigbee devices with Homey as well?
Thread & zigbee use the same chip ( called MultiPAN).
If true, consider using one protocol @ Homey and the other with HA for instance. This can reduce the mesh issues (still “chatty” devices without a setting to limit the events traffic should be replaced)

But none of these issues affected my setup before 13.3. I’ve only used stable releases since late 2024.
I’ve had those Eve and Aqara sockets for the best part of a year and there has been no delay in flows and no devices dropping off, until I updated to 13.3
In fact, I actively encouraged other users to buy the Eve plugs, because they made my mesh stronger, more reliable and with faster response times for all end devices.

I have thick brick walls in a UK period property. I have always needed multiple Thread routers or nothing works.

I actually removed ALL six of the Eve sockets on Wednesday night and the issues I have described persisted and haven’t got any better

The ignorance that the firmware isn’t to blame when other people describe the same issues, is quite staggering to be honest.

Like deejayreissue said, I haven’t had any issues with my setup before firmware 13.3.
My setup:

  • only matter devices (like 10 over WiFi which are fine) and 20-25 Matter over Thread
  • no ZigBee etc
  • Firms: Eve / Aqara / Govee / Nanoleaf
  • no extra apps installed besides Anker for Solar / Shelly / Matter Bridge (which I’ve uninstalled to test if the problem persists - it does)
  • HomePod Minis and Apple TV with Thread as Border Router
  • Homey in the same Thread Network as Router / Homeassistant, too, tested with Discovery
  • some Devices with Thread firmware 1.4 and some with 1.3

Devices are first and newly installed over Homey (Homey over iOS), and after a while in Apple Home (iOS Keychain) and Home Assistant.

Devices work reliable in Apple Home and Homeassistant. They stay online. It’s the same network, IPv6 ist working, mDNS ist working etc. - and it worked in Homey before the last firmware update.

Now in Homey it can be, that all Thread-only devices are ok and after a while they are getting connection lost - and come back randomly. I have Automations in Homey and it’s a bit of luck if they are working whilst the devices are there or have a connection lost.

I have no logical explanation


they come and go

Could you please create a diagnostic report and reach out to our support team? They’ll be happy to help you investigate this.

Just to make sure I understand correctly: you’re using the Apple Thread network, and Homey is joined to that same Thread network?

If the devices remain online in Apple Home and Home Assistant but go offline in Homey, that could indeed point to an issue on our side. It’s difficult to troubleshoot without diagnostic data though, so if you send a diagnostic report to our support team, they’ll be able to take a closer look :grinning_face_with_smiling_eyes:

Correct! Apple Thread and all Border Routers in the same network. I’ll try to do so asap

There’s a difference between causation and correlation :wink:

You mentioned that you already removed the Eve plugs, but based on the other two brands in your setup and their Thread performance, I’d start by taking a closer look at the IKEA devices. They’ve earned a bit of a reputation when it comes to Thread reliability.

At the moment, we haven’t received any other reports of people experiencing exactly what you’re seeing. The threads you tagged me in are about Matter-related issues, but they’re different scenarios with different setups.

Also keep in mind that we’re talking about a platform with close to a million users. It’s not unusual for two people to run into somewhat similar issues around the same time, even if the underlying cause turns out to be completely different.

Did you know that you can also connect those via the Govee app (regular cloud connection, no Matter)? This will probably work more reliable than using Matter.

Patronising me in the same thread as someone with what appears to be the same issue, is quite the flex