Ezlo Atom specifications


FCC information about the Ezlo Atom is here: https://fccid.io/2AMY9EZLO100

The embedded processor is a 32-bits RISC processor (Tensilica L106), which can go up to 160MHz.

1 Like

I am familiar with the ESP8266 SOC… It is the low cost 2.4GHz wifi which has made all the wifi outlet plugs so cheap and popular. This is not a replacement for the edge/plus/secure. It is a new low cost wifi-zwave bridge with very little processing power and memory. Is most of the processing going to be in the cloud?


If the Atom is supposed to run the in-house RTOS then it might be fast enough. There’s a lot of overhead for the current OpenWRT model. But I do share your concern.

1 Like

The internal photos show the Flash being a Winbond 25Q128, which is 128M and makes more sense.

The processor is ESP32-D0WDQ6, which is dual core, no Flash (makes sense, uses external) package with WiFi b/g/n (150Mbps) and BT/BLE. Datasheet says 520KB SRAM, and that’s… not much. It also claims 600MIPS, but in a RISC processor that’s a pretty spongy statistic. But that RAM… there’s one chip on the board I can’t identify, and I’m hoping it’s an SRAM.

How will this compare to the 2008-era MIPS 1004Kc in the Plus? I guess we’re going to find out. But as @HSD99 said, a dedicated RTOS may make all the difference, at least for the core software.

Oh, and thanks to the fellow who submitted this package to the FCC and scrubbed all the chip labels from the CPU-side internal photos… except on one photo. You da real MVP. :smiley:


If you want better pics of the inside:

1 Like

Interesting… this has different chips from the 4/18/19 FCC submission. Your submission is dated March 2018, so I’m guessing the acquisition changed the chip specs a bit. A couple of parts on the newer submission definitely look hand-soldered (the Q6 processor package especially looks really sloppy, but I doubt I could do better, that’s too small for my aging eyes and hands).


I have high hopes for their RTOS and independence of cloud processing, otherwise my guess is there will be many Atoms running Tasmota or ESPEasy for a cheap wi-fi/z-wave chip with local control.


Even if Ezlo said to the FCC that “Ezlo Atom” and “ZL-100” were identicals, it’s possible that the “new” Atom will be slightly different (at least we know that the outer parts are different), but probably not very different.

The platform is probably fine in order to run “native” handling on an efficient RTOS, but we are also used to a web interface, plugins and scripts. For the web interface, even by pushing most of the processing on the client side (in a similar way to AltUI), it will already be an interesting challenge. What is even more challenging is scripts and plugins. My guess is that most of the handling we are doing through plugins and scripts would have to be directly done by the native handling, as scripts and plugins on such a platform would be very very challenging. This is consistent with Ezlo looking to be very interested in knowing which plugins are considered to be the most important ones by current Vera users: if the new platform doesn’t have the luxiry to run plugins, the required features must then be handled “natively”.