Raspberry Pi Zero 2 W
raspberrypi.comI've been testing one for a couple weeks, and a quick rundown on it: it has a quad-core A53, same CPU/SoC core as the Pi 3, with 512 MB of RAM.
This means it can run 64-bit Pi OS, but it is memory-limited and even some of the benchmarks I ran needed extra swap space to complete. But for anything that does run in the 512 MB of LPDDR2 RAM, it is at least twice as fast as the Zero/Zero W, and sometimes 3-8x faster, depending on how things handle threading (4 cores is better than 1!).
My biggest concern is overheating—even at the base 1 GHz clock, it will start to throttle after a few minutes in a case with no airflow or heat sink. You can overclock to 1.2, 1.4, or even in some cases 1.5 GHz, but it starts throttling a lot more quickly unless you have really good cooling.
Also adding on from another thread that got demoted from the front page: I tried usb boot on this board (since it has the same SoC as the Pi 3 B, which supports USB boot), but alas, that didn't seem to work.
The Pi Zero 2 W doesn't seem to do anything at all (no LED activity, no HDMI output) unless there's a microSD card inserted with a valid bootloader and DTB that supports it.
Definitely seems a bit disappointing that 5-6 years after the original Pi Zero there's still only 512MB of RAM. Regardless, I'm excited to see what new projects will become feasible with the extra CPU performance.
I just finished watching your YouTube video, so thanks for providing all of this insight. A new Pi Zero will certainly make small handheld emulators a lot more promising...
I don't feel like I needed more memory or the cpu boost in the pi zero 2. I'd rather they stayed at the old $5 price, decreased power consumption, and added some analog and realtime capabilities. I had thought that was the idea of the RP2440 chip used in the pi pico. It would be great if they added that chip to a pi zero board (maybe in the same SOC as the main processor) so they would have some analog inputs and a somewhat realtime coprocessor.
I've done some fairly fancy Python app development under Linux on a 300mhz embedded ARMv6 board with 64MB of ram, so for lots of things the Pi Zero is already overkill. But an actual OS (i.e. Linux) does make programming a lot easier than an MCU board with an RTOS. So how about a "Pi -1" (minus one) with 1/4 of the resources and power consumption of the Pi 0?
Power savings is complicated. A Pi board is going to have a minimum power usage with the CPU only being a portion. A lower clocked CPU on a Pi -1 might reduce peak power usage but might not get much savings at idle vs a comparably configured Zero. The Zero would also have the option of underclocking for peak power savings with the ability to spin back up to work faster to return to idle faster. Don't forget disabling board components can yield baseline power savings on Pis [0].
[0] https://mlagerberg.gitbooks.io/raspberry-pi/content/5.1-powe...
I think you'd need a SoC designed for phones instead of general computing to reduce power by that much.
Also, I'm curious what use-case do you have that would need an RTOS?
512mb probably means the power budget it's designed for doesn't get completely stomped.
If you need more memory just get the pi4 and a suitable power brick for it
According to the launch post, they also couldn't find a single-die 1 GB RAM module, so they couldn't physically install more unless they upped the complexity of the design significantly.
Do you have power consumption numbers? Compared to the Pi Zero W, hom much more does it consumes?
Yes; check my blog post [1] or latest YouTube video [2] for details, but the tldr is 0.4W (80 mA) idle minimum for Zero W, 0.6W (100 mA) idle minimum for Zero W 2.
Zero W goes up to 160 mA at load, while Zero W 2 gets up to 400-500 mA due to the three extra CPU cores.
[1] https://www.jeffgeerling.com/blog/2021/look-inside-raspberry...
That's a big jump. Is there a way of getting Linux to disable a couple of cores?
`echo 0 > /sys/devices/system/cpu/cpu[0-n]/online` should do what you want.
Even as root, that gives `permission denied` on Pi OS on the Zero 2 W. I'm trying the kernel cmdline option too.
Edit: That works! See: https://www.jeffgeerling.com/blog/2021/disabling-cores-reduc...
Well… I said it should, not that it would. ;-)
On x86 there was an option, i think nr_cpus or something like that some time ago. There was also SMP_KERNEL kernel config option.
2W! Did you measure Gflops? I wish you could disable cores dynamically...
That's why they called it Pi Zero 2 W and not Pi Zero W 2, which is what it is.
> I wish you could disable cores dynamically...
Another comment says /sys/devices/system/cpu/cpu[0-n]/online doesn't work, but I think you can get pretty far by changing the CPU affinity at the cgroup or process levels (for userspace) and on /proc/irq/default_smp_affinity (for interrupts). There are likely some kernel threads that would still run on the other cores, but this is reasonably close.
In particular, looks like setting CPUAffinity= /etc/systemd/system.conf will set a default for all of userspace that you can override for particular things as desired.
Is undervolting an option?
As for the RAM limitation, zswap is a great way to get some more RAM at cost of some CPU cycles which on the new Pi won't be an issue.
I've been running some tests on my regular server, where it seems that zstd offers better compression ratios than the default lzo-rle. Ratios are currently at 3.7 vs 2.6 according to netdata; I've seen them as far apart as 4 vs 2. Mostly a matrix server with a big database.
zstd could be slightly slower, although according to some benchmarks, it decompresses faster.
zram*
zram is the compress-in-memory like you have in Windows and ChromeOS and zswap is the compress-on-disk in case you want to replace SD cards.
zswap does not compress-on-disk either (I wish it did). The difference is just an implementation detail into how both plug into the kernel but the goal is the same on both.
zram is a compressed ram-disk that acts like any other block device, so you can point `mkswap` and `swapon` at it.
zswap is a compressed (memory) cache that uses the frontswap API to evict least recently used pages to a swap device. Evicted pages are not compressed.
zram is more of a general purpose component; whereas zswap is a more specialized component.
https://www.kernel.org/doc/Documentation/blockdev/zram.txt https://www.kernel.org/doc/html/latest/vm/zswap.html
64bit is great news (if only because I can target it with the same binaries), but only 512MB of RAM and no EMMC is disappointing, regardless of power budget constraints.
I have been using 3As because of the reduced size (and the full sized HDMI port, which comes in handy), so I hope those get updated to a 4-grade chipset —- I feel that there is a gap between the Compute Module and the A that needs filling.
next we need a RPi4A with 2GB of ram, the full size HDMI, usb C and updated silicon
At some point we have to ask what the raspberry pi platform is for - they used to be low budget, low cost, low power general purpose computing in an extremely small formfactor.
It seems like there's still one product, but increasing divergent use cases for that product.
I'm not entirely sure what you're saying, but it sounds like you're arguing against their strategy of having multiple variants of the Raspberry Pi for sale at the same time?
There have been a lot of commenters in this post basically complaining that there are too many Raspberry Pis. From a practical standpoint, I don't get the complaint (I mean, how does that personally harm them in any way?).
The Raspberry Pi foundation's mission is to increase technical literacy worldwide by making and selling very low-cost yet reasonably high-performance computers. They also do quite a lot of education and outreach themselves.
What people here seem to be forgetting is that these things take money and instead of spending time fundraising (a huge cost and time sink for most non-profits), they decided to tap into the maker and computer geek communities, as well as industry. When people like you and me buy these boards to play with or build into our projects, we are subsidizing all of the foundations design, education, outreach, activities. Also when companies build products and internal solutions around the Pis. The whole reason they still sell the (mostly) original Raspberry Pi board is because there are systems, processes, or fleets still using them. They'll continue to use them until a design change is required. As long as its relatively easy for the foundation to squeeze out a new batch every so often (all of the hard work of designing them is already done), what does it hurt?
I don't think we need to ask that at all. They make little computers for people to tinker and do interesting (or even boring) things with. The "diverging use cases" are a feature, not a bug.
That might well be the sweet spot, yes.
Does it overclock/throttle automatically like Intel? E.g. if I put some good cooling, will it run on faster frequencies?
I could get it to run stable at 1.2 GHz in open air. 1.4 required good forced air cooling and still was a little unstable. Someone got it to run at 1.5 GHz but it was a bit unstable at that clock (overvolt of 8).
Inside a case, it quickly heats up under load!
In all cases, the CPU frequency starts stepping down when the temp hits 80C, and on the default clock it will go to 800 MHz until the temp is 79C. If it reaches 85C, it will throttle severely, all the way down to 400 or 600 MHz, I believe, until the temps get back down to 84C.
have you tested lz4? it's probably a little worse ratio, but it is ridiculously fast to decompress.
Is it 28nm or 40nm?
I believe it's 28mm like the Pi 3 B.
Original RPi0 supposedly can idle at 0.4W, where-as the 3 (whose cpu this is based on) idles at >1.5W. It's cool to have this great small package, but this is a very different offering than the first RPi0W.
It's felt really weird seeing so little progress in available low power ARM over the years. Atmel/Microships SAMA5D3 / SAMA5D2 with it's 600 MHz Cortex A5 (2009) still seems like- half a decade latter- one of a rare few genuinely low power Linux running SoC available for purchase/use. I'm forgetting what's available in the Cortex A7 line, which is a little higher power but which still has some very low power chips available for it, & at significantly higher performance ranges. A7 has a lot of chips you can't buy, only available in some tablets & or hard to get wearable chips, but I think there are some good offerings about too. ST has the ~$20 STM32MP1, to name one.
Yet I just see so much non-materialization, non-availability. It feels like the rule. 2015's Cortex A35, 2016's Cortex A32, 2019's Cortex-A34... none of these chips seem to have really materialized in any genuinely available form. There are some off-brand tablets with A35 chips but generally that's about it. The low power segment has been rather ignored.
The processors on the Pis aren't really optimized for low power, and they aren't strictly ARM. These Broadcom chips have a VideoCore that starts up first, which then brings up the ARM cores (geerling goes into greater detail here [0]). I don't think there was any kind of work done to properly suspend both cores for sleep modes.
If you're looking for low power ARMs, maybe other boards may fit the bill better.
The Pi 3 model A/B have a lot more electronics—check my video and blog post about the Pi for more on power efficiency—the Zero 2 W can run at 0.6W (100 mA) vs the Zero W at 0.4W (80 mA).
Both idle a little higher if you're actively using HDMI and WiFi.
i.mx8 Nano supposedly can idle at <0.2W, with max power somewhere around 2W. They have also announced ULP version if i.mx but that seems somewhat vaporware at this point.
As for the relative unpopularity of Cortex A3x series, I suspect that might partly because Cortex M series has been creeping into similar performance territory, like for example the i.mx rt mcu on Teensy 4.x, which has Cortex M7 running at 600 MHz.
"Low power" and "something that runs an OS as heavy as Linux" are two conflicting features.
I've been running one of these for a few months too and was excited when I saw an ARMv7 CPU. In the open source and container world, many projects have stopped supporting the Zero and RPi1 due to its lack of RAM and slow CPU speed. Even running "go build" can take multiple seconds on a simple program. Starting Node.js was also fairly slow.
The RPi zero shines for Python scripts that access hardware - for sensors and camera projects that are low power and cheap to make. It was actually perfect for the GrowLab project that we did over the summer with 20 others from the community https://growlab.dev - sensor data and camera images were aggregated to a more powerful RPi3/4 and then either uploaded to a static GitHub Pages site, fed into InfluxDB to create beautiful charts with Grafana.
The new Zero 2 means that I can start to run ARMv7 or even ARM64 containers on the zero again, but with the limit of 512GB of RAM. The launch blog post explains why this couldn't be made higher.
The first thing I tried out was not K3s, which I knew already suffers on a Raspberry Pi 3, but faasd. Faasd is OpenFaaS but built for pure containerd, no multi-node networking and no Kubernetes. It works fairly well for a few functions, even with NATS and Prometheus being deployed as part of the stack.
So whilst you're not going to be building K3s clusters with these, they can run containers - with Docker, containerd, nerdctl and even as a full application stack with faasd.
Where this gets more interesting for me, is hosting small applications, integrations or APIs. Perhaps with Ingress via a tunnel that can penetrate NAT/firewalls like Inlets, Argo or Ngrok.
I've ordered one of the release models to see how it performs. Look out for a blog post from me soon.
512GB should be enough for most people ;-)
MB
Used to be kb. How times have changed.
Shoving 128kb into my 6502 build right now like "that oughta do it"
There isn’t much 6502 software that can use more than that.
My //e had (technically still has) 320K. Most of the time, it was a ramdisk.
Sure, but I'm the one writing software for this computer and I do /not/ trust myself to be efficient.
Of course I haven't tried banking in anything over 48KB yet so I don't even know if my circuit works, but it boots!
found the same experience last night, k3s to heavy but I have faasd going now. just have to figure what to do with it now
Surprised nobody has mentioned or compared it to the Radxa Zero.[0][1]
Spec of the Radxa Zero ($15):
CPU: Quad Cortex-A53 1.8 GHz, 12nm process
GPU: Mali G31 MP2
RAM: LPDDR4 512MB/1GB/2GB/4GB
Storage: eMMC 5.1 8/16/32/64/128GB and uSD card
HDMI: Micro HDMI, HDMI 2.1, 4K@60 HDR
Multimedia: H265/VP9 decode 4Kx2K@60
Wireless: WiFi4/BT4 or WiFi5/BT5
USB: One USB 2.0 Type C OTG, one USB 3.0 Type C host
GPIO: 40Pin GPIO, ADC/UART/SPI/PWM
Others: Crypto Engine, support external antenna, one button
CPU: Broadcom BCM2710A1, quad-core 64-bit SoC (Arm Cortex-A53 @ 1GHz)
RAM: 512MB LPDDR2 SDRAM
Storage: None, MicroSD card slot
HDMI: Mini HDMI port
Multimedia: H.264, MPEG-4 decode (1080p30); H.264 encode (1080p30)
Wireless: 2.4GHz IEEE 802.11b/g/n wireless LAN, Bluetooth 4.2, BLE
USB: 1 × USB 2.0 interface with OTG
GPIO: HAT-compatible 40 pin I/O header footprint
Other: OpenGL ES 1.1, 2.0 graphics, CSI-2 camera connector,
composite video and reset pin solder points
> The Radxa seems to be better value.
The Radxa is a better value if pure performance is the only criteria.
But the real value of the Raspberry Pi products is the ecosystem and code support. It will be much easier to find tutorials and software support for the Pi Zero 2 W than the Radxa and the Pi will be supported for a long time after the Radxa has been discontinued.
> But the real value of the Raspberry Pi products is the ecosystem and code support.
Also in availability. I just clicked buy and will have a Raspberry Pi Zero 2 W shipping to me in two days. I looked on seeedstudio for Radxa products; literally nothing is in stock.
That's an interesting setup, I'd love to hear more. Do you have a wifi antenna hooked up to the pi?
As client radio I use a cheap rtl8812au-based single-antenna Wi-Fi USB stick by Piaek with a ALFA APA-M05 7dBi antenna, and as access point I use the on-board radio of the RPi in master mode. I use iptables to set up NAT (masquerading), optionally via an OpenVPN tunnel. It can be an odyssey finding a cheap Wi-Fi stick for which there are good drivers available (Amazon reviews and answers do help), let alone one that can be run in master mode, and you often have to compile the drivers yourself, in my case (Arch Linux ARM) I need to pull the dkms package from AUR, change the PKGBUILD to build for Pi and it builds awfully slowly (around 1 hour).
As for long-range antennas, you can check out Yagi's or Cantenna, also very nice DIY projects. The longest range can be achieved by dishes. Though check your countries dBi limit that is allowed to be emitted in any direction and then reduce your tx power to stay below that. The directivity will still help reducing interference from other directions. I like the APA-M05 for its compactness.
For the cooling, I use DHT22s and simply run a bunch of fans via a PWM-driven MOSFET, directly from the Pi, roughly following the video: https://www.youtube.com/watch?v=oJ32CMxliCQ
I send notifications (actually, not SMS) via qpush.me.
Yay, another thing i won't be able to find anywhere for a year, then find it for $25 with "only one per customer"and $15 shipping.
I understand it's a global chip shortage, but when zero W came out, there wasn't one.
It is currently available (and in stock) from the official resellers, at least here in Europe. Limited to one per order but that kind of makes sense considering the global chip shortage. Get one while you can. :)
Good news: in stock including US shipping at https://shop.pimoroni.com/products/raspberry-pi-zero-2-w for less than $25. Probably cheaper from US resellers but I'll leave that to you. Yes, one per customer, which is a reasonable policy right now.
Bought one this morning from PiStore.us for shipment to the United States for $15 + $8 shipping.[1]
Pi Zeroes always looked like a scam from Canada, as the approved resellers limit you to 1 unit and shipping was more expensive than the unit. Now at 15USD, it's the first time shipping is cheaper than the unit, but I can't get it below 35CAD (28USD) shipped to Toronto.
Super exciting! We use the previous gen at my company to run octoprint software for our 3d printer farm and they do tend to run quite slow. This should be a nice needed upgrade.
In general the RPI Zero series is really great value for a tiny linux machine.
According to the OctoPrint Twitter handle, it runs great as an OctoPrint server[1]!
[1] https://twitter.com/OctoPrint3D/status/1453608862188249095
I can't wait for this too. I've got a Prusa printer that's got the perfect spot for Pi Zero, but its never quite had enough juice to run octopi with all the trimmings.
side rant: it's annoying how you can't buy these things by themselves, usually bundled with crap
I do want this, I think it's taken for granted how easy this thing can connect to a camera. I bought a Teensy/blue/BeagleBone/etc.. thinking "Oh I'll just plug in a camera into it". For the Teensy it won't but maybe the Beagle bone. I was using a single core Pi Zero so it was noticeable doing things like iterating over frames. Granted it can just be poor code on my part/noob in OpenCV.
It seems pretty easy to order just the Pi? I opened the US suppliers, and of the ones that aren't already out of stock, you can:
buy directly from Makerbright: https://makerbright.com/raspberry-pi-zero-2-w-.html
or place an order, which will ship in early November, from:
* Sparkfun https://www.sparkfun.com/products/18713
* PiShop https://www.pishop.us/product/raspberry-pi-zero-w-2/
* CanaKit https://www.canakit.com/raspberry-pi-zero-2-w.html
Zero WH and Zero bundles are unlimited but bare Zero/Zero W are limited to 1 per customer/account/shipping address. Allegedly sold at loss, that’s also why Zero WH exists.
Thanks for that, yeah I saw the two on the rpi site suggested were already out of stock. I do normally buy from Amazon/locker but I just grabbed one from MakerBright.
Part of me feels bad whenever I use these so much computing wasted, like I have one taking a picture of an indoor garden/turning a light on/off with a cron job/Apache server and that's it. Oh well, good time for tech anyway.
Ah thanks! I also just realized other vendors would likely ship later. Wish they mentioned them in the official blogpost as well.
I think that's more due to the 3rd party seller than the raspberry pi foundation. I just picked one up from the pihut[0] without anything bundled in.
Random thought upon reading the linked announcement:
If you stacked multiple Pi boards on top of each other, connected through the GPIO HAT connector, could you network them?
Apologies, just woke up and coffee not kicked in. This is probably silly :)
EDIT: yes, I should search for this sort of thing before posting: https://forums.raspberrypi.com/viewtopic.php?t=127660
tl;dr "serial port, RS485 + SLIP of some sort, probably would max out at around 1Mbps, 4Mbps theoretically possible with small MTU's" and "There is no hardware reason why an RS485 network could not be used as the physical layer for at least 32 nodes, and with some of the RS485 driver chips, 127 nodes (they have stronger drivers)"
RS485 is just an electrical standard for serial transmission that includes multiple devices on the same bus. It is not a network or transport protocol. I doubt SLIP would function on an RS485 connection as 485 is typically used in master-slave configurations where only one master unit sends messages and listening "slave" devices only respond if addressed. You'd have to write a protocol layer to handle the transport of packets over 485, even in four wire full duplex setups.
The issue with "stacking" is the pins on each board are not part of a bus but individual programmable io pins. Any stacking would require an intermediary board to isolate all these pins safely while adding the necessary TTL serial to 485 interface chips (e.g. Maxim max485).
CAN is similar to 485 where it's a 2 wire bus though it has smarter hardware which enables full duplex communication and any device can listen or transmit. Unfortunately it is pretty slow compared to Ethernet but plenty fast for its intended use.
If the RPi SoC supported being an SPI slave (it doesn't) you could do something similar to a token ring network at up to - theoretically - something like 100 mbit/s.
or just use the standard rs485 protocol that all industrial hardware has used for decades and has preexisting nice libraries - modbus. Built in CRC, dirt simple, reliable.
I know modbus very well. The OP was talking about networking the machines using SLIP over 485 which is completely different than moving around bits and words.
One thing which I miss on RPis, especially these tiny ones, is a tty exposed via the charging USB.
I'm not sure how that works? you connect the pi to your laptop and use putty to login?
This would also require one of those tty-to-usb cables?
No, just a regular USB cable. USB's CDC protocol makes it look like a serial port to the host PC.
Or even better USB OTG.
You can do this with gadget mode, the g_serial module.
Sure, but it has two USB sockets of which one is only used for power supply; if that one could also be used as OTG (instead of just TTL RX/TX as suggested by the fellow) that would be great.
How long before cutting edge hardware trickles down to stuff like the Pi Zero? I think a usb-c port would be more useful for these small form-factors, but it's probably a long way of considering there's only usb 2.0.
Yep lack of a USB-C power port is a no-go from my perspective. I only have a handful of micro usb cables left in the house, not interested in adding more at this point. Even the cheap terrible LED mini projectors, and high end Kindle e-readers have all switched over to USB-C for power. Strange to see micro usb on a Brand New Product in Q4 2021.
Apparently this has it: https://forum.radxa.com/t/introduce-the-radxa-zero/6550
I ordered one, then quickly cancelled because of the lack of pre-soldered header. Looking for a drop-in replacement for my Pi Zero WH due to time constraints (new baby).
I presume they didn't make the pre-soldered header available at launch due to high demand and the ongoing chip shortage. Much easier to produce one less complex package.
Despite having just 512MB RAM, I think it's perfect for the its common use cases. For me, the CPU increase will be helpful for getting it set up and installing software (especially if it needs to be compiled), not so much for when it's actually being used.
too bad, no 5ghz wifi.
Yeah, this was by far my biggest disappointment with it. The Wifi on the old one was a constant problem for me, and this one doesn't seem to improve it.
Tangential but important : supply :)
Here in Australia, seems like it is available from Nov 8 at Core electronics [0]
[0]https://core-electronics.com.au/raspberry-pi-zero-2-w-wirele...
When you're talking about the Zero, supply isn't a tangent. There has never been a strong enough supply of them.
It's great that finally it no longer use armv6.
Now the Pi ClusterHAT can go to 20 cores of pure arm64 performance!
Welp, excited to not be able to buy one of these.
One thing that's been interesting is that despite the 2x markup of anything pi related where I live, the pi pico was available almost immediately at about $6. I guess I'll be going a bit lower level then.
> a micro USB port, and a micro USB power port
That's pretty sad.
I can think of a half dozen reasons to justify going with micro B but I feel the same.