🬤 rns.recipes

Forum / Help / Thread / Export

# Raspberry Pi Zero 1 W

_Help · started by Anonymous on Mon, Sep 14, 2026 9:19 AM_

Tags: Question

---

## Original post

**Anonymous** · Mon, Sep 14, 2026 9:19 AM

Used, they are very cheap to get and also use very little power. Do you think they are strong enough for a small rnode plus local wifi setup? I did not see anybody using them yet, so I am unsure if its a bad idea for some reason.

---

## Reply 1

**Memo** · Mon, Sep 14, 2026 3:13 PM

Last year I used a **Raspberry Pi Zero 2W** for a sensor node. That was really pushing the limits. I rather doubt that a Raspberry Pi Zero W will work.


---

## Reply 2

**Emru** · Mon, Sep 14, 2026 4:16 PM

I don't see why it shouldn't. Question isn't "would it work", but rather "how fast will it work". I think I have RPi0W hidden somewhere in my junk, also I have RNode, I could set up RNS on it and check it.

Also, I've found this

https://github.com/rfjakob/gocryptfs/issues/452

There is a benchmark for RPi1, but Zero uses the same SoC. For AES stream, throughput is 4.8 MB/s. ZeroW uses higher chip frequency (1GHz instead of original 700MHz), so it could be even faster. Of course there are other factors, but I think it could be feasible as bridge between RNode and WiFi, or as some sensor data provider.

If someone wants to run RNS on SBC, I'd advise to use aarch64 architecture, as it has AES acceleration in the specification. Most chips provides it... except those used in RPi3 and Rpi4 :). RPi5, RK35xx, and other chips should support accelerated AES.

https://gpages.juszkiewicz.com.pl/arm-socs-table/arm-socs.html

In this table, check for 'aes' and 'sha2'.

As I can see, Reticulum uses AES from Cryptography, which in turn I assume uses OpenSSL, which should use hardware acceleration. As for SHA, in code there is an import of hashlib, so it depends if Python uses hardware acceleration, and since when.

Check Orange Pi Zero 3, it's more expensive than RPi0W, but a lot more powerfull, and has all those hardware support. I also have it, so I can even do benchmark between them.

Have fun :D

---

## Reply 3

**Anonymous** · Mon, Sep 14, 2026 5:33 PM

Interesting! Exactly the type of information I was looking for, thank you. If you happen to get some actual data, please share it. I will start doing some research myself :)

I found the raspi zero 1 for as little as 15€ per device. So thats quite a lot cheaper then a new orange pi zero 3.

---

## Reply 4

**Anonymous** · Mon, Sep 14, 2026 5:35 PM

**Memo** wrote:

> Last year I used a **Raspberry Pi Zero 2W** for a sensor node. That was really pushing the limits. I rather doubt that a Raspberry Pi Zero W will work.

>

> [image]

Did you measure it, or did it just feel sluggish, and in what way? I am. nterested what pushing the limit actually means and if its something I maybe am ok with.

---

## Reply 5

**Anonymous** · Mon, Sep 14, 2026 7:22 PM

Just thinking about this. My local Lora parameters create a bandwidth of around only 1kbps. Looking at the AES benchmark from before it feels to me that the lora bandwidth will always be the bottleneck however slow the raspi zero might be. Which would mean it would make a good Rnode base.

---