That's because RaspberryPis are no longer cheap throw away computers meant for education or hobbyists, they're developer kits for manufacturers that need a CPU running a well supported mainline Linux in their products.
The only reason they don't cost $500 or more is because the foundation needs the hobbyist market to write and support the open source BSP, without which the RPi would be just another poorly supported also-ran in an already crowded market. With how well supported mainline Linux is on the Pi, EEs would be willing to pay a lot more.
I think a lot of people don't realize that there was a decent size, but small market for SBCs for low-volume embedded work (including hobbyists) before Raspberry Pi. You could get a lot of different kinds of boards with good Linux support and a not terrible price. Often, a processor vendor would explicitly provide support for these things because it was a good vector for selling chips.
Broadcom, having never wanted anything to do with this market since volumes were too low, had an abundance of a CPU SKU that was good for this. So some broadcom engineers founded Raspberry Pi to use up this excess stock, essentially getting these chips for free. The original RPi blew every other SBC out of the water on price performance (and many manufacturers out of the market) because by getting the most expensive component for free, they could sell Pis for an extremely low price. It also massively expanded the market for SBCs, as hobbyists flooded in to work with RPis.
5-10 years ago, the sweetheart deal with Broadcom went away. Now Raspberry Pi has to compete with everyone else for Broadcom SoCs, and during the semiconductor shortage of 2020, Broadcom had tremendous leverage. Now, Raspberry Pi pricing is nothing special, but they still have the brand name and they have captured the community (on the back of behavior that was borderline anticompetitive).
This is spot on. There is a market for small SBCs that spans from weapons to washing machines. When the RPi was introduced it threw a huge wrench into that market because it came in, with an operating system and storage, at under the price for the typical enclosure, much less the board itself of the existing systems. Look at PC104 systems for example.
The two pieces that have to be in place, as 'threshold' requirements are
1) The SBC exists and is available from a manufacturer
2) The same manufacturer provides an OS for that board and its associated board support package (BSP) which is drivers for all the I/O and system support functions.
The industry is full of people who went out of business because they chose Vendor A's SBC and Vendor B's OS, only to fail to deliver when it didn't work with Vendor A and Vendor B point at the other saying it was their problem. So people just don't do that any more.
What most vendors in the SBC space, prior to the introduction of the Raspberry Pi, didn't have was 20 to 30 thousand programmers writing random bits of code. What that meant was the Pi's feature set exploded rapidly, what's more there were lots of free tutorials on programming it.
In the SBC space before Pi that "Programmer Training" was one of the ways the vendor made better margins at $500/hr for a class of "up to 15 students" at our facilities.
So before, higher priced SBC + BSP, and you had to send your programmers on a road trip to the vendors facility to get the hands on training, and then you had to pay every time you made a service request.
After, cheap SBC + BSP!, a bunch of different programming videos on the web for free! Program doesn't work? Just ask the community of enthusiasts what they think!
We are not surprised a whole lot of the smaller SBC vendors closed down after that.
Yup. That is where the money is, it isn't in "hobbyists".
It will be interesting to see if the RISC-V efforts will allow for vendors to supply "50 year" SBCs (which is to say they are guaranteed to have pin-for-pin and bug-for-bug replacement SBCs for the period of 50 years.) That was a weapon system requirement back in the day that got waived for electronics because vendors couldn't "force" chip makers to keep making the chips they would need for repairs when there was no volume. But if you can define a socket/footprint and then drop in an FPGA "core" which can evolve but always appears to the circuit to be the same processor, that will be an interesting evolution point.
Another thing is that Arduino is focusing on is going for the mechanical engineers. This is a very lucrative market for electronics products. Just ask National Instruments and every other company that makes a PLC.
Sometimes I think NI blew it by not maturing LabView into something stronger. I've been on way too many factory floors that have their QA/Burnin rigs controlled by a rats nest of LabView modules.
> I think a lot of people don't realize that there was a decent size, but small market for SBCs for low-volume embedded work (including hobbyists) before Raspberry Pi. You could get a lot of different kinds of boards with good Linux support and a not terrible price. Often, a processor vendor would explicitly provide support for these things because it was a good vector for selling chips.
Before RPi became really popular and the 4 came out, I used BeagleBoard, PandaBoard, Aria G25, Gumstix, Cubieboard, i.MX devkits, and spun custom boards using Marvell, TI, Freescale (now NXP), and Qualcomm CPUs - I don't remember any of them having as good a BSP or being as easy to develop with as the RPi was five years ago. Maybe my memory is (very) faulty but the experience was leagues worse. PTSD-inducing level of worse. I wasted weeks or months on every major project shaving silicon yaks that should have been handled by the vendor (and is now handled by the RPi community).
The modern i.MX toolchain may now be comparable many years later but I've long since given up on everything else since CM4 came out in 2020.
My understanding is that they lost the sweetheart deal after they pivoted to supporting commercial, right in time for the pandemic supply crunch.
Yocto works the same way whether it’s Pi or iMX, most of the learning curve has nothing to do with the SoC. So it’s really strange to hear that your Pi workflow is better than anything you’d get with another chip..
Two products.. leveraged a Yocto-based board support package (BSP); one being in AgTech, and the other being in the veterinary space. These products have followed disparate practices when leveraging the BSP for custom hardware and software.. this talk [described] the two products, how the BSP was customized and used, and the resulting consequences.*
Yeah, one is easier and the other is right. Shipping a reproducibly built readonly rootfs image takes longer, but is strictly better than putting 'some' versions of Debian packages on the SD card and calling it a day.
It's the Arduino curse - sure, it's faster to ship Hello World with Arduino but soon you realize all your libraries were built by beginners and use delay() everywhere so you're screwed if you need two peripherals to work at once.
You use Pi for prototypes and one-offs, not where you need to ship something that's actually competitive on BoM.
No, your memory is spot on. I also eval'd dozens of ARM (and some x86) SBCs for embedded use back in the early/mid-2010s and most of the BSPs were awful. You'd be locked into some ancient several-year old Linux kernels and documentation (and even actual hardware support) was often buggy/incomplete. God help you if you need to customize the boot/spin your own (Yocto was still new then, I assume life is a bit easier now).
> I think a lot of people don't realize that there was a decent size, but small market for SBCs for low-volume embedded work (including hobbyists) before Raspberry Pi. You could get a lot of different kinds of boards with good Linux support and a not terrible price.
And you still can! The big innovation from Raspberry Pi was making it all feel very accessible through the documentation, community, and various utilities to configure things via menus instead of by editing files.
The Raspberry Pi was rarely the best board, but it was the easiest to recommend to beginners because you could point them to volumes of documentation and community threads.
Joining the queue behind the histories of the subsequently retconned sole credit for the RPI's design, and the termination of E Upton as the Foundation's CEO.
While I agree... it has been much better supported than the alternatives. I've tried several other SBCs and they have all had issues that I didn't experience with the RPi. Not that RPi hasn't had issues of its' own. Namely in power requirements, the USB-C in physical but not electronic sense for power, to overloading most USB power supplies in practical use in Pi 4B models.
I switched to the cheaper Intel mini pcs when RPi supplies were short, and scalping made it much more expensive. A whole PC with more ram, case, psu and faster storage that was faster for under $200 vs $150+ for an 8GB RPi 4 board only was a no brainer.
> by getting the most expensive component for free
I get that you're exaggerating, and you perhaps aren't trying to be deceptive by misusing the word "free," but the low markup the RPi guys initially paid to Broadcom does not explain as much as you think it does about the Pi's success. To explain why let's examine something else you're wrong about:
> You could get a lot of different kinds of boards with good Linux support and a not terrible price. Often, a processor vendor would explicitly provide support for these things because it was a good vector for selling chips.
Prior to the Pi you could get one board with good Linux support and a not terrible price from a vendor who provided support because they thought it was a good vector for selling chips. That was the BeagleBoard, from TI. I mean, the BB barely checked all those boxes: the support wasn't very good, it was kind of nonexistent compared to the support community the Raspberry Pi people created. But they sure didn't have to worry about the cost of the CPU.
So back to the original point: getting the CPU "for free" (since we're apparently just saying "free" now when we mean "at cost") wasn't a decisive advantage for the Raspberry Pi people, since the TI people had the same advantage. TI had a few other advantages as well, like a first mover advantage, their own assembly lines, and relationships with everyone who sells electronics components.
TI's support was crap when compared to something like RPi which deliberately targeted newbies, and as far as I remember they didn't have a few amazing people in their community dedicated to making a whole new spin of Debian and supporting it like RPi did. And, you know, PR matters. All that stuff is what made the difference.
> anticompetitive
I guess we're just throwing words around without caring what they mean today for some reason.
> I get that you're exaggerating, and you perhaps aren't trying to be deceptive by misusing the word "free," but the low markup the RPi guys initially paid to Broadcom does not explain as much as you think it does about the Pi's success.
Broadcom sold the chips to the Raspberry Pi organization at a significant loss, not at a low markup or at cost. They were not free, but they were close to free. The TI guys never even gave anything close to "at cost" to the BeagleBoard folks.
Also, the BeagleBoard was the most hobbyist-targeted and the one with the best PR. There were, and still are, tens of companies making SBCs, but before the RPi they would all sell you singe unit quantity. Not any more. Most of them actually had better support than TI. The NXP boards in general were and still are my favorites.
Also, providing a product at a loss so that it significantly undercuts your competition (also called "dumping") is very much anticompetitive. I don't like using the term for Raspberry Pi because they clearly weren't out to create a monopoly, but the Raspberry Pi was dumping a product.
> Broadcom sold the chips to the Raspberry Pi organization at a significant loss, not at a low markup or at cost. They were not free, but they were close to free. The TI guys never even gave anything close to "at cost" to the BeagleBoard folks.
Everything in that paragraph is “citation needed,” unfortunately, but I certainly don’t mean that in a negative way. I wasn’t aware that Broadcom lost money on the early Pi parts and I feel a little skepticism about that. You seem to know something about what TI charged the Beaglebone people, I’m curious about that as well.
It’s not that dumping parts to establish a competitive advantage is beneath Broadcom - what’s beneath Broadcom? It just seems rather… prescient of them? Unless you were talking about the Pi 3 generation or something. I’d be more inclined to believe that.
On the point of PR, I don’t remember Beagleboard having anything comparable to Pi’s seemingly organic enthusiasm twelve or thirteen years ago. But I guess I’m not sure I would remember.
That’s not all meant to be some kind of RPi love letter. They were great at all that community building stuff, and in my view the best at it, but in light of the IPO it’s quite a joke.
Is there any reliable source for this? It's an interesting claim, but I'm skeptical because I don't think anyone had any idea what a huge success RPi would be. The idea that it was all a monopoly play by Broadcom is something that I'd need more evidence to believe.
It's not a monopoly play by broadcom originally. It was a way to get rid of excess stock of a chip that was going obsolete (without just tossing them) while doing something good for computer engineering education. I don't think anyone expected this to get so big.
Subsequent chips were specially made for the Pi by broadcom, and supposedly they didn't have as large of margins as other customers.
Is there evidence that Broadcom sold those chips to the foundation for less than their own cost? That’s the claim I find a little strong. And I do understand that undercutting other nonprofit or small vendors might also be considered bad behavior, but it never sounded like there was a lot of demand for those particular chips (which were already on the edge of obsolescence, and were even in the process of being abandoned by Debian).
I had very much the same feeling. Not because Broadcom are nice guys, but because it just doesn't sync with what everyone thought they knew at the time.
> you perhaps aren't trying to be deceptive by misusing the word "free," but the low markup the RPi guys
You perhaps aren't trying to be deceptive by misusing the word "markup," but citation please for /any/ "markup the RPi guys initially paid to Broadcom"
Huh. If I knew what their markup was, I would not have asked pclmulqdq for more information about what their markup was. Or what their markup wasn't, if pclmulqdq is right and Broadcom gave them the chips at a "significant loss."
Yes, as the article discusses, there were also the options of just using a mini-tower or SFF desktop, or a NUC/booksize PC, and often these options were comparable or less in price.
For example, I owned an ECS Liva 2/32 that I got for $100 AR and a pair of ECS Liva X 2/32 that I got for $125 each AR. These used standard N2807 and N2808 Bay Trail-D processors (Silvermont), meaning they are OoO, run standard x86 binaries and distributions, come with storage onboard (which was far more reliable than the early days of fullsize SD cards), a proper adapter, a case, and USB 3.0, for essentially the same price you paid for a Raspberry Pi 1B once you factored in all the little extras.
Obviously if you want to do GPIO and stuff, the RPi is a better option, but that also competes with micros, as mentioned. And a lot of people ended up using it for some variety of non-GPIO IOT thing attaching some USB device to wifi/ethernet, or as a NAS, etc, and the Pi was terrible at these. It had massive reliability problems (unthinkable today) and I went from incredibly excited to bouncing off it hard and learning a lesson about the right tool for the job, and I think I wasn't alone.
Part of the problem was the decision to use USB 2.0 as a system bus. Everything including network and disk all hung off a single half-duplex USB 2.0 high-speed bus. And the Rpi firmware had a bug in the USB stack which dropped frames under load, so actually this was unreliable and could lead to corruption all on its own. It took over 2 years after launch for the foundation to fix this fundamental bug in the literal system bus of their hardware (and thanks to broadcom's closed documentation/blobs, they were the only ones who could see the info to fix it).
On top of that, people forget about the non-micro-SD card slot. The fullsize slot just let the card cantilever out into space... but the thermoplastic used to make SD cards actually will soften enough from the heat of the rpi over time to warp away from the contacts and lose connection. Even just stepping up to eMMC (not the most reliable) was a massive step upwards because of how bad the Pi was at not melting its SD cards.
And the power adapter situation has always been dire - most USB chargers were utterly unable to provide consistent enough power to avoid brownouts and corruption, and (while you could PXE boot) the hardware was not even capable of PXE booting without at least a helper SD card to load from. Today, it's dire once again, with the Pi 5 drawing 5A @ 5V from the adapter, which essentially no adapter on the market is capable of delivering.
The 2B and later models (especially 4B) are enormously faster, the move to OoO cores instead of literal a single-core A57 or whatever improved performance a ton. The change to USB 3.0 (5gbps full-duplex vs 480mbps half-duplex) and getting the USB frame drop bug fixed helped a ton (the pi couldn't even serve a usb hard drive at full speed). And the 4B moved to armv8 which has a drastically clearer standard for binaries than the armv7 situation, which was an utter mess at the start (softfp binaries were a thing for a bit with the early 1B releases iirc!). Admittedly Pi was better than any other arm device but the Liva just ran debian or ubuntu.
I actually think it was an utter failure as a product. I cannot imagine having to maintain a whole computer lab of these things melting (softening+warping) the SD cards, can't even PXE boot unassisted, can't get good adapters? Let alone the situation with firmware problems/driver quality and needing (at the time) weird binaries with a non-standardized bootloader etc. Why would you buy that instead of the Liva X for the same finished price?
The pivot to "these are for makers!" was a pivot, and then they pivoted again into commercial. Like good for them but it's always been a messy product with an uncertain target market.
It was always borne largely on the backs of enthusiasts shoving it into various home-server and IOT usage, most of which didn't actually need a GPIO. And the GPIO thing was always better served by the ESP8266 in most situations, and that ecosystem really matured almost in parallel with the Rpi. So sure, while the Pi got more suitable for actual GPIO hobbyism over time, so did everything else too. I regard the whole thing as fundamentally having been overhyped and a waste of time by people who (in most circumstances) really should just have bought a booksize or an off-lease OEM USFF desktop.
That was me. Got an old Core2Duo desktop for $10 at a surplus sale, that became the fileserver I was trying to build, which instantly ended the reliability struggle I'd been having (rpi lasted an average of 1-2 months MTBF in my homeserver usage) then I got a couple booksize and nettops and ended up with NUCs etc which are just a far better fit for what I'm trying to do.
To each their own, but again, I think really very few people are interested in them for the GPIO stuff, it's the "$35 for a little computer" thing that draws people in, but that's an impedence mismatch to expectations, that's not why you should buy a rpi/if that's your use-case then there are better options.
It's late and you may never see this but I do appreciate this post. It also reflects my own experience. The Pi2 still awful (I never encountered the first version) and I eventually put it in the electronics recycling bin.
I also like your perspective on the bigger picture here.
The Model 1B was $35 (in 2012 dollars) and the still available Model 4B starts at $35? It might even be argued that the Model 1B's successor is more the Raspberry Pi Zero 2W for $15, which is cheaper than the original.
The Raspberry Pi 5's base model does start at $60 but its specs are too different for a comparison to be meaningful.
[EDIT] Oops, I hadn't realized the 4B with 1 GB was discontinued. So the starting price of the 4B would be $45 for the 2 GB version.
There are complete passively cooled n3350 based systems on Ali express with 64gb storage and 6gb ram in a case with power supply ready to go for $65 with free shipping. That works out cheaper than the cheapest pi after buying case, storage and power for the pi. You can buy usb gpio breakouts for <$10 too. Lower power than the pi 4 too due to the huge process node advantage (despite the x86 disadvantages). 28nm vs 14nm for the pi 4 vs the n3350.
The pi is fun but honestly for pi hole or similar you might as well buy the all in one x86. For media streaming definitely buy the all in one x86. For gpio stuff ok the pi is reasonable but even then if you want to make a product rather than a home automation once off you’d go a different route completely.
I suspect a lot of the current disdain is a product of function creep. While the original Raspberry Pi was used as a desktop and server, people understood its limitations. Now that many of the limitations have diminished, to the point where you can expect reasonable performance as a desktop and use it as something more than a simple web server, people are justifiably comparing it to alternatives (which have come down in price over the same period of time).
Of course, the Pi is also facing competition from higher end microcontroller based solutions. People seem to forget that there was a time when hobbyists bought the Pi for "Internet of Things" like projects, both due to its cost and size. Then came the ESP8266 and ESP32 and development boards that packaged both a microcontroller and network interface.
This, but also that they gave a huge middle finger to the hobbyist community during the component crisis by giving preference to integrators.
Most makers I know have pivoted to ESP32 during this time as it was good enough and actually available. Probably would have happened sooner or later though.
The ESP32 has nothing to do with a Raspberry Pi. It's only good that people learnt to use the better tool (in terms of price and power consumption) once better tools like MicroPython or NodeMCU came around.
When the first raspberry was introduced, it was really really hard to interface an electronics project with the internet. Arduinos were really dumb at the time. That's why the raspberry was so ideal. But most electronics projects didn't really need a whole linux distro running on it. It was just that there was no other option.
The ESPs introduced a totally new class that can cover most of the usecases of electronics projects that the raspberry originally aimed at. They support wifi, bluetooth, pretty serious processing, enough for most connected projects. They totally ate the raspberry's lunch in the embedded market. Of course the RP2040 aims for this too but IMO it's kinda too little too late, the ESP32 is already so well established and has the biggest community.
At the other end of course the PC pricing came down and the intel N100's and the like eat its lunchon the other side.
Absolutely right... I used to use Pi Zero W's for IOT projects. I had several of them around my house running sensors, rgb lights and the like. Now I can do the same thing with a much smaller device like an ESP32. Setting up a device that changes light colour based on time can be done with a €3 esp8266 flashed with WLED. For other projects I will use the beefier ESP32-S2/S3 or the Pi Pico Microcontroller.
If you have a led to light up over the internet the Pi was the simplest, cheapest and fastest to develop for. Now the ESP is simple enough and so cheap the RPi doesn’t make sense for the use case, but it took a long while to get here.
Too true.. I have a nas project in pieces on the table next to me, since I was out of work for several months and couldn't put any more into it... I "saved" a hundred dollars or so going from AliExpress components, the mainboard doesn't work right, and trying to get it replaced would cost about as much as a different board from AliExpress. In the end, I wish I'd spent the extra hundred or so on a different case and board. The cost of the drives significantly outweigh other costs.
It bugs me to no end how much a small (4-drive) nas can cost, compared to much more capable mini-pc prices.
That’s all true but the OP said that Pis are “no longer cheap”. The reply was simply a demonstration that they are still available at the same price point, no matter what the competition is or isn’t doing.
Sure, if you want to deal with strange problems no one has ever faced. RPi's strength isn't in meaningless performance benchmarks, is in actually getting stuff done.
They vary widely. I bought an off brand mini PC where the USB ports randomly lost power. There was no online community or support to speak of to help me figure out the problem. In that regard the Pi is far better.
Are there loads of driver support for those USB/GPIO things? I've only done that on PI and the Python libs made it super easy. Now it's one more thing to solve research rather built-in.
I don't think so - if you needed GPIO on a small x86 the easiest way would be to hook an Arduino / RP2040 to it. That seems like it's still the sweet spot for RPi, esp. the Zero W, if you need small, low-power, full OS and GPIO.
Not really. I've been researching this extensively lately to try to add GPIO to my stack of Dell Wyse 3040s.
Options with mainline linux kernel drivers:
- MCP2221A (i2c + 4x GPIO)
- CP2112 (i2c + GPIO)
Options without kernel support:
- FT232H
- Arduino nano and clones
- Raspberry Pi Pico running some interesting firmware
- ESP32 running something like ESPHome (completely separate from host)
I've chosen the Pico for now and forked the u2if project for firmware and host support[1]. I also put together a generic ESPHome-compatible protocol server in python to tie my widgets to Home Assistant[2].
Why do everyone latch onto bit-banging pins from Linux??? Isn't that going to exhibit wildly unpredictable delays? Wouldn't you normally compartmentalize hardware handling into an OS-free or hard-realtime microcontroller and let the uC communicate with host CPU in less-realtime manners?
You can roll your own "RGB temperature sensing modem" and get/set values with `bash` and `expect` over ttyUSB, no one PWM controls RGB LEDs on a gaming keyboard straight from a task tray app. Trying to wrestle GPIO problems that way is going to be unnecessarily hard.
Not sure which part you're referring to, but I agree. The u2if firmware I'm using has an interface type that generates the addressable LED bitstream with the RP2040's PIOs. The host software dumps the pixel color values over USB CDC into a buffer on the RP2040, then when it's done the firmware tells the PIO to read the buffer with DMA and stream out the precisely timed LED bitstream.
Currently it's running in "one light string per PIO" mode but there's an inaccessible mode where one PIO can do up to 8 strings simultaneously. The pins have to be physically adjacent, though, so it's a bit more difficult to set up in a generalized way.
Ah, yeah, that’s the right way to do it, but if you can get away with using a Pi you don’t have to write a second program for your Arduino, haha. People who are happy to do it the right way were already well served, right?
Yep, that one is pretty cool too. It only does one i2c bus so it seems like it underuses the hardware a bit, but the mainline driver is pretty valuable.
What kind of bolted-down expansion does a cheap n3350 box have? How fast can I swap the main storage for something completely different? Can I power it with my USB battery bank? Does it support two digital monitors or does it instead have one HDMI and one VGA (like this is 1987)? Can I use it like an appliance and just plug in easy-to-download images like LibreELEC or EmulationStation, or do I need to understand how to make computers work before I can have a good experience with Kodi or console emulation?
I can see plenty of n3350 systems with dual hdmi on google and plenty with drive bays and relevant connectivity. Those seem to cost ~$100 vs the minimalist $65 ones.
In terms of driver support these intel systems are the easiest to work with in Linux. Very mainstream and well established drivers.
There are some nice cases for RPi 4 that have a semi-integrated USB-SATA adapter in the box. The DeskPi[1] case, Argon and the NESPi-style case in particular. DeskPi is pretty neat, if costly, and the NES-like cases have easy 2.5" sata swap. YMMV.
I upgraded my home LAN to 10gb/2.5gb early last year, and the RPi can't really keep up with it anymore. I'm using an N305 based mini-pc with 4 ethernet ports for my router now... exceeds my needs and runs great. My home server is an AMD 5900HX based mini-pc with also works very well (ProxMox, Docker, etc).
The only USB thing it has is a NIC. But given the speed of my ISP I could get around that by using VLANs, I guess.
I mean: There's not much to it. LAN port, WAN port, MicroSD storage, and a (completely optional and of dubious merit) hardware RTC that sits on the GPIO header. I'm using the bog-standard Pi Foundation plastic case.
It runs OpenWRT, which itself is very light-weight -- it's more-or-less just enough Linux to set up routing functions using a GUI, and it's meant to be able to fit in the tiny little bit of flash that consumer router hardware comes with.
IIRC, the status at the time when I put it together was that it could route packets and perform traffic shaping at ~1Gbps even with a USB NIC. It certainly did just fine when I had a 400Mbps pipe.
Inflation adjusted the 1B (with 512MB, which was the only available variant) was slightly more expensive than the 5 with 2GB at their respective release. The price "increase" is that 4 and 8GB models are available at all and most people buy them, presumably because people actually don't care about the purchase price so much. But you still get the cheapest variant, if you really want to. Also, there are different variants of the Zero, all of which are cheaper than the 1A at release.
The power brick got more expensive due to the increased power use. Also if you want to make use of the full power, you need a cooler. But you can also do without cooler, because it just gracefully slows down when overheated.
I think the Pi drove the price of lower end computers down so far that people completely lost their sense of perspective.
I was about to make a quip then I realized it’s actually somewhat of a valid point: if we’re going to inflation adjust USD purchase price, we need to inflation adjust RAM requirements. 512 MiB in 2012 went much, much further than how far 512 MiB will take you in 2024 (especially for the low-end desktop usage, rather than the embedded cli users).
While this is true... you reach a point where it's perfectly reasonable to compare a mini-pc to the higher end RPi when you consider the full cost to build with power, storage and enclosure with price/performance.
> Oops, I hadn't realized the 4B with 1 GB was discontinued.
Is it? I can buy one for $35 at my local Microcenter, and I can see that Adafruit is selling them for $35 as well. [0] It's still listed as a $35 part on the official Raspberry Pi product page. [1]
I checked Digi-Key through RPiLocator and they report several thousand units in stock each of the Raspberry Pi 4 2GB at $45 and of the Raspberry Pi 3B+ 1GB at $35: https://rpilocator.com/?vendor=digikeyus
For those unfamiliar with them, Digi-Key is a electronic components supplier to manufacturers but they also sell to individuals. Their stock count should be accurate.
Look harder? They're available at a number of retailers, and if you're in the US, Adafruit is recommended - but I wouldn't pay more than $35 + shipping in any case. There are a couple of dozen online retailers here with the 4B 1GB:
I see the Raspberry Pi model B+ available right now on Adafruit for $30. It has a single core 700 MHz CPU and 512M RAM.
The Raspberry Pi Zero 2 W is available right now for $15 with a 1 GHz 64 bit quad core CPU and 512 MB RAM.
So it seems to me that you can get a Raspberry Pi SBC today, that has higher or equal specs in every regard, for a lower cost than the original. Am I missing something?
Looking at https://github.com/raspberrypi/linux, I scrolled a few pages of commits and didn't even see one from outside the Raspberry Pi org. I'm picking through their merged PRs and it looks like maybe there are a couple, it's hard to tell.
But it looks to me like they just do a great job supporting their own product?
> So it seems to me that you can get a Raspberry Pi SBC today, that has higher or equal specs in every regard, for a lower cost than the original. Am I missing something?
It’s just supporting whoever needs a RPi 1st generation. It’s the same reason some older models of things cost the same or more, it’s basically fishing for the “we need the exact same hardware” customer.
The Pi Zero models are also kind of priced to move to make a statement of “look how much power we can stuff in a $5 board”; IIRC they were at first sold at a small loss to hit that $5 price point.
The Pi 5 is still not supported mainline. Proper mainline support for older models was contributed by third-parties, not the RPi foundation, which just care about their kernel fork.
> they're developer kits for manufacturers that need a CPU running a well supported mainline Linux in their products.
Raspberry Pi 5 isn’t as well supported in mainline. You’re still going to be using their kernel if you want all the features, just like many other module these days.
> The only reason they don't cost $500 or more
$500 is a huge exaggeration. There are numerous modules and small boards well under that price that come with good support, including many with full x86-64 CPUs.
I think it’s more correct to say that the boards are approaching equilibrium with other boards and modules in price, not that they’re secretly some premium $500 product sold at a discount for reasons. Nobody would be buying Raspberry Pi anything at $250, let alone $500.
Wasn't there a very long time that the Pi wasn't fully supported on mainline? And it's boot sequence is still a bit weird in that the GPU handles bootstrapping?
I think the main point is that you know the company/product support isn't just going to disappear into the sunset in 2 years and you're stuck with an increasingly outdated hacked together kernel thrown over a wall
Given that Raspberry Pi hasn’t actually been well supported by mainline Linux and that the Raspberry Pi foundation hasn’t put a lot of effort into upstreaming things, I don’t think it’s actually a big deal. People don’t care where their kernel comes from as long as it works.
I am surprised by all of the comments here that assume Raspberry Pi has great upstream support. It’s amazing that people just assumed their boards were working great with upstream kernels. Raspberry Pi has a history of doing nonstandard things that serve their community but are actually a little bit quirky when it comes to normal embedded Linux.
I highly doubt so. In fact, save for the RP2040 which isn't Linux capable (0), all their processors aren't for sale anywhere; Broadcom simply won't sell them to you, no matter if you order 1 or 100000. That is, you can't build your product around one of their CPU and you have to put their entire boards in you product instead, which translates in huge costs, no industrial rated parts and forced use of SD cards for system disks, which in that context are a no-no. The RPi is a hobbyist board with a huge potential for teaching, but I wouldn't consider it for anything beyond that use.
0: yeah, I know you can run it in theory; I mean in a usable way.
That forces you to sandwich two boards together (connector, more costs, etc), and still you have no freedom regarding which peripherals to place around the RPi SoC.
The average integrator doesn't want to mess around with things like DDR5 routing or designing power supplies for the SoC. There are lots of other companies such as AMD and Nvidia selling SOMs so the fact that they are selling SOMs can hardly be something to complain about.
Plenty of PIs in labs of hardware tech companies. Very convenient when you need a small linux box, indeed, or when you need to access a piece of equipment over a serial terminal, et.c
The pi has a lot of exposed pins and associated hardware capabilty. That was an intrinsic part of it's design. It's what got me interested in electronics again. Any comparison should include that. It was never meant to be just a computer.
This. The Pi is a great little real world <-> computer interface. The hat ecosystem is really cool. Using one as a general computer is foolish. Unfortunately each Pi generation seems to move more towards the computer case and away from the real-world one.
It really is! I built a system to monitor/control my wood stove with a pi. The GPIO pins made it easy-ish to hook up an infrared temp sensor over i2c. I found a relay hat that allowed me to switch 120 VAC to control the stove's fan. My pi 4B had plenty of power to run a Ruby on Rails server to build little charts/graphs, a config UI, and some cron jobs. Lots of very powerful Lego blocks, basically.
I bought a few used NUCs for $150 each, they're amazing home servers. Much, much faster, more capable, more flexible than a Pi, at only twice the price.
Pi's were never a good for compute-intensive tasks; running a DNS resolver (PiHole), print-server, or uploading backups to the cloud is the sort of thing it excels at. The Pi Zero W costs $15 and idles at less than 1 Watt; it's the always-on-appliance that turns on my GPU-workstation or NUC for the heavier stuff.
What do you use them for? I'm one of those that have Raspberry Pi 5 mostly just for Home Assistant. It's clearly overpowered for that use case, but I wanted the NVMe support. I'm not convinced by these articles -- used x86 box that maybe achieves almost the same low power draw still doesn't have any actual upsides if you don't realistically need that computing power.
I have one where I deployed K9s so I can learn Kubernetes better, and one where I have deployed Harbormaster (http://harbormaster.readthedocs.io/).
The Harbormaster one has a bunch of stuff (Zigbee2MQTT, my smart home stuff, my apps, etc. I have a Pi 4 that has Octoprint, services on the NUC load instantly whereas Octoprint feels a bit sluggish.
The NUC is an x86 (well, amd64) box, with a 10W power draw, which is great. I don't think a desktop PC will do less than 100W...
I use the mini-pc for running security NVR, git servers, self hosted NextCloud, local hosted LLM models, and backup servers.
After my branded security camera NVR died for the N times and the company went out of business, I got fed up and decided to run my own NVR. I got a powered switch to provide power and network connection to all the cameras via POE. Connected a mini-pc to the switch and ran Blue Iris on it to stream from all the cameras. Attached an external harddisk enclosure to use some old harddisks for storage. This ensures every piece of the system can be replaced and upgraded separately.
I have a home server that stores and processes all my photos, using PhotoPrism. Things like face recognition does require a bit of a compute. A NUC is perfect for this use case.
I've got one that does PhotoPrism + other media (sabnzbd, gerbera, flexget) as a general "media storage" box, one that just runs a Minecraft server, and one that's "everything else" (currently Home Assistant, Grafana, Prometheus, my webcam bird detection stuff, NATS, etc.)
Originally started with HA on a Pi 4 but it wasn't really up to it.
Just as a quick update after trying out: seems like a very nice local photo album service. The initial scan will take a long while, but looks like Pi5 is quite enough to handle the service after that completes. I will need a larger SSD for my Pi if I intend to keep all my photos in this, though.
I bought a 8TB SATA SSD for this. Currently it's already 15% full. The nice thing about my NUC is that it has Thunderbolt builtin so I can buy Thunderbolt drive bays.
So you're saying that the Pi is a proverbial Toyota Hilux and people are buying it to highway commute to work and back instead of using it for what it was meant for, and then saying it compares poorly to sedan? Yea, no shit.
If you're not using the GPIO or any of the ribbon cable peripherals there are much better options out there. But a Pi will be able to do things those machines never will.
I think they are comparing a $700 micro PC to an $80 Raspberry pi, but pretending they are in the same price range since the micro PC is available cheap in the refurb market.
You don’t need to pay that much. For example, Minisforum is selling a barebones MS-01 for $399 new. This isn’t quite apples to apples — the Raspberry Pi includes RAM (but not much), whereas the MS-01 includes a case, a cooling system, a power supply, and an RTC battery. (And the MS-01 uses a non-janky 19V supply, whereas the Raspberry Pi 5 wants a weird not-to-spec not-quite-sure-what-they-were-thinking 5V USB supply by default.)
For the price, you get massively more CPU power, 3x the number of easily connectable NVMe devices, at higher bandwidth each, 22x (!) the network bandwidth, and the ability to connect a real multi-lane PCIe card of your choice.
I still find it sad that NVMe is an afterthought in the Raspberry Pi ecosystem. SD is convenient, but it’s also slooooow, and it holds back a lot of raspberry pi use cases. The new-to-RPi5 official NVMe support still seems really awkward in the way it interferes with the overall thermal performance and the way it interferes with the IO header if you use the official board.
You don’t need to pay that much. For example, Minisforum is selling a barebones MS-01 for $399 new
That doesn't really change my point that they are comparing the Pi to a PC that costs 4 - 8 times more (and in a much larger formfactor), so it's not surprising that it's faster.
Personally, I'm a fan of the odroid H+ series - full x86 SOC with dual nics, and some on-board gpu stuff for $125. I have 3 of the old H2 series and I love them.
Not only are they very different in price, the micro PC is 10x the size. Size doesn't matter for all uses, but there are going to be things where a Pi will fit but these micro PCs won't.
I got one recently that was £159 (1.4x the 8GB Pi 5 starter kit) for a 16GB i3-8109U (with 500GB SSD) and is just about 3x the volume of the Argon Neo 5 case for the Pi 5 (113 x 127 x 43mm vs 94 x 70 x 30mm).
Doesn't change the "[places] where a Pi will fit but these micro PCs won't" assertion but the pricing and sizing are not as egregious as you're suggesting.
You can pick up something like an HP elitedesk g2 with 8th gen intel cpu for $100-120 on secondary markets. Most of them ship with at least 8gb ram and also a 500gb m.2 drive. A 8th gen cpu will be many many times more powerful than even the rpi 5. In addition to having quicksync for hardware media transcoding. Something the Pi still cannot do via hardware. Not to mention, its x86 - so lot more support over the ARM7
Of course, if you size and the ability to interface with other hardware via GPIO is the primary use - then yeah a SBC like a Rpi would be the better option. But for a small home server, one is better off just buying a mini-PC.
Now the 4020 isn't certainly a monster, but I can assure you it's way more performant than a RPi. Also, bear in mind that, as is the case with many Chinese products, those mini PCs are produced in huge quantities and sold under at least a dozen different ever changing "brands". Don't let the name "Wo-We" make you think this is something deemed to disappear in a few months; the name could certainly be thrown away but the product will most certainly reincarnate under another "brand".
OK I understand the RPi must have very good press, but nitpicking every part of a message just to find something to attack isn't constructive. Linux preinstalled means that Linux works out of the box, that is, you don't even have to search around for Linux compatibility with any of the peripherals inside that mini PC. Of course I would never ever trust anything preinstalled, neither Linux nor Windows, just as I got rid of stock Android on my recently bought Pixel 7 in favor of GrapheneOS like 2 hours after unboxing it. I only meant about compatibility, certainly I wasn't suggesting that anyone runs unknown software from China.
You misunderstood. My comment had nothing to do with the meat of your post. I was only referring to Linux being preinstalled on Chinese hardware as being downright dangerous.
I got the $700 figuring by looking up the list price of one of the mini PC's mentioned in the article, which use an Intel i5-6500 or AMD Ryzen 3 PRO 2200GE CPU (which each have 4 cores @ 3.6Ghz), I think if they'd compared against a 2 core 2.8Ghz Celeron, it would have been more evenly matched. The Raspberry Pi 5 beats the 4020 in both single and multi-core performance in this benchmark:
Don't let the name "Wo-We" make you think this is something deemed to disappear in a few months; the name could certainly be thrown away but the product will most certainly reincarnate under another "brand".
Does that matter? If it dies in 2 months and I can't find the manufacturer because it's operating under a different name now, is that any different than if the manufacturer folded completely?
I'm not sure whether to assume you either are clueless or whether to assume you're actively working to try to sabotage perception.
No one's gonna recommend an ancient middling mini-PC that costs vastly more than a modern current mini-PC.
Either go on the secondary market & get this old PC for under $100. Or go get some modern Ryzen mini-PC for $500. Or a decent n100 for somewhere in between, especially if you insist on new for some reason.
The manufacturer is the same, what dies is the brand name; the very same product is just being packaged in a box with another name. I mean, you shouldn't place too much importance in the name, just look at the real iron inside the box.
We, as westerners, are used to give a lot of importance to brand names, possibly because it comes from the old times when brand names identified products with the families that created them; that is completely different from how it works in far east today.
The reason I give importance to the brand name is because if the company has been around for a decade (or many decades), it's likely to still be around in a year or two to give me support when the product fails.
If the company resurrects itself every 6 months under a different name, then there's effectively no warranty on the product and there's no reason to think that it's built to be long lasting, since even bad reviews won't show up under the new brand name.
For $150 you can get a new mini pc with a low power intel cpu (e.g. N100) and 8GB of Ram. It comes with an SSD, a power supply and (gasp) a case. Add those to a raspberry pi and the price is not much different.
I have been looking at a N100, are they good in general as a Pi alternative. I am mostly looking at running K8s and running apps and running things like Pihole
I recently got a mini PC with an N100, 16GB ram, 512 GB SSD for €150, and am using it to run several docker compose projects. Basic home lab apps, such as nginx, home assistant, adguard home, vaultwarden (bitwarden), and actual and the CPU is at 4% and RAM is at 22%. I don't know how intensive K8s is, but I'm very happy with my setup and am surprised it's not breaking a sweat running all these things.
They're all over Aliexpress (prices before VAT, annoyingly) and often posted to hotukdeals when there's a sale. The Ryzen based miniPCs are also pretty good. Here's an example.
I think these are great options for where pis aren't needed. Over the years I've seen a deluge of "I needed a microcontroller but didn't know what that was so I used a pi" and "I needed little more than a docker image but I didn't know what that was so I used a pi". The pi really comes in handy when you need the combo of the 2. Otherwise, people are just jacking up the price for those who really do need it. And that's not me, but rather people I've known who had great use cases and couldn't buy them.
After having bought Pis and then sold them all, I've never understood them. The Pico and Pi Zero seem to have a place, but the performance of the big Pi is so bad, it's rather pointless as an "embedded" computer or general purpose computer with a display.
For an embedded computer you basically need to go bare metal with Circle or something similar.
But then I'd wonder what you're building because there are powerful microcontrollers you can buy for $15/1 that will handle anything with basic networking and sensors. I know some musical synthesizers are made with rPi4 and I'm befuddled that they're not the most powerful synths ever made.
I think they oddest one out is the Arduino line, which is generally underpowered and expensive compared to just having a drawer of esp32s sitting around.
Arduino pre-dates the existence of ESP, at least in the western market.
Also, Arduino as I think you're using it here is really just slang for AVR microcontroller dev boards.
Arduino isn't actually that, it's a boot loader and a highly simplified set of libraries to interact with a wide variety of microcontrollers including ESP-32 and what people traditionally think of as "Arduino" meaning the branded dev boards labelled that way.
Of course the whole Arduino ecosystem is basically garbage, but it does help beginners get into the idea of doing embedded things.
>Of course the whole Arduino ecosystem is basically garbage, but it does help beginners get into the idea of doing embedded things.
I think this is either a naive take or a presumptuous, Gatekeeping take. The Arduino ecosystem is not garbage. It’s an abstraction layer. While it is not always the most efficient code, it makes programming something straightforward. That is not only a beginner feature.
The premise that’s sold here is that because Arduino is an abstraction layer and it does (often) favor ease over efficiency and compactness, it must not be suitable for “real embedded programmers”
Not all projects need to be accomplished in 256 bytes of flash. Not all programmers care about how to set a given timer or stepping down clock speed. Not all use cases are your use cases
Use the tools that get the job done. Make the thing you want to make. Ignore people who tell you you’re doing it “wrong”.
The Arduino ecosystem is objectively garbage, for a host of reasons. Yes, it is an "abstraction layer", that attempts to abstract away all the hard stuff from a whole host of different microcontrollers, and does it in some really bad (and broken) ways.
Arduino prioritizes making things easy for the person who doesn't know how to write embedded code. That's fine, as far as it goes. The problem with things like this is when people put them in production anyway, because "it works" on the bench and there's no need to bother with making sure it actually works.
I agree with this in general. I think the Arduino ecosystem makes a lot of sense and as mentioned upthread does a good job lowering the barrier to entry.
It's the hardware I don't get, which sits in a weird spot in the matrix of ease and power.
A embedded computer using microSD for main storage should be a non starter for any serious application. They fail far too easily given the bad thermal layout on the board. You can get the larger ones to boot USB, but the smaller ones obviously can't.
I've got a guy who loves running these things, but calls me every other month because one of his images fails, and he needs help rebuilding it. So far I doubt he's saved himself any effort, time or money.
This is so true. Back when I used to use reddit, I had to leave the raspberry pi subreddit for this reason. 95% of the projects only needed a small C program and microcontroller but instead used a full blown OS and Python. It drove me nuts.
I just wholeheartedly disagree relative to Arduino. You're telling me I gotta find a microsd card, get their software, either
A) muck around with the boot settings file (though I think recently they can do that in the flasher) or
B) find a mouse, keyboard, monitor
Then, as a beginner who likely doesn't already, learn Linux and beyond that embedded Linux, then learn the languages and environment (the amount of times people have told me their task was to get the project's code started on boot is astounding), networking, drivers, etc etc etc.
You're telling me that's 1000x more approachable than the 1 line of
digitalWrite(BUILTIN_LED, 1); // because Arduino will set the output mode for you I believe
It takes on an Arduino? Which boots immediately every time, no sd card corruption issues, no anything, just runs the code?
Nice overview, another venue worth looking at is thin clients. Some models like the Fujitsu Futro s740 are passive cooled, draw only 3-4w idle, can encode hevc in hardware and support up to 16gb memory and a nvme drive. There is a nice overview here: https://github.com/R3NE07/Futro-S740/blob/main/README_EN.md
I've been using some of these off-brand n100 mini PCs as a homelab cluster for the past year or so.
Their physical size is smaller than a Raspberry Pi with case, and performance is more than twice a RPi5. And you get full x86 software compatibility. Idle power is around 4-5W measured from the wall.
I use the odroid H series(1) for basic 'small server' usage. Dual nics for use as a firewall, DDR5 upto 48G, multiple sata ports, m.2 port, etc. Totally silent and very low power draw. They run from like $125 to $175 new, depending on model.
I've had 3 of the old H2 series and I really love them...
Also worth considering getting the $6 Pi Pico microcontroller in such a pairing. Keeps the microcontroller capabilities at a lower cost and without having to maintain 2 operating systems across 2 different architectures.
Yeah, any PC can have GPIOs if you plug a cheap USB microcontroller into it. That's more or less how the Pi5 works internally anyway, as they've moved the main SOC to more modern silicon processes its internal GPIOs have become less able to tolerate hobbyist abuse, so now they proxy the GPIOs through their custom southbridge chip instead, which is an amalgamation of a microcontroller and various other peripherals.
Indeed, a USB MCU is my preferred GPIO these days. It makes my peripherals platform independent, and I can do my code development at my comfy desktop workstation with its big displays.
I find it easier to write real-time code on the MCU.
In fact, I've disciplined myself to make all of my projects -- hardware and software -- capable of running on any modern platform. It turns out that's not hard to do.
There are so many USB GPIO modules that I don't think the Pi is so "obvious." Plus if you blow out your USB GPIO, replacement is a far easier thing to consier.
How does the idle and loaded power draw compare between an Rpi 4 (or 5, if you prefer) and an N100? Genuinely would like to know, because I think power usage (and heat) is an important part of the equation.
Very much this. If you just need a small, low power computer to be a server or whatever, it's hard to beat used business/enterprise SFF computers or a new N100 based NUC.
rPi's defining use case is as a microcontroller, it can also serve as just a computer but it most definitely isn't optimized for that.
I don't see the appeal of using these mini-PCs. Using an old second hand power supply is enough to turn me away.
And the PI has many advantages. Power supply dies? I can order a new one in literally seconds. And meanwhile I wait for it to arrive I can use a spare notebook charger or whatever. Etc.
The shortage sucked, but it is solved now.
Last image I created I tucked in a Raspberry Pi 1, and it worked just fine. The versatility is unmatched. Equally I can test an image at my home and then let someone install it on the other side of the globe.
The point of a Pi for me is more that you can have it where you need it. Attached to your TV or whatever.
For a home server I would recommend something beefier, an old desktop will be superior to any mini-pc and the Pi. But probably bulkier and more power hungry.
A Pi5 with nvme does work and be a decent home server for tinkering though. From my perspective the niche for the mini-PC is pretty much nonexistent.
But I don't think maximum utility is the goal though, it is a hobby. Do what you enjoy! Tinkering with low power PCs might be enough of a reason alone.
But these comparisons to the Pi doesn't make much sense to me.
> Power supply dies? I can order a new one in literally seconds.
For the RPi 5, for fully reliable operation, you need an oddball 5V/5A power supply, which is not actually a standard device. IMO Raspberry Pi messed this one up. At the price point, either use a barrel jack or support USB-PD for real. (The latter would be great for many use cases, because the same conversion circuitry would enable driving from a wide voltage range. A tiny cheap ESP board can do this — why not a rather pricey Raspberry Pi?)
I can’t find any evidence that RPi5 supports PPS or that anyone makes a 5V/5A PPS charger, except perhaps a massive 100W charger or something along those lines. It’s a silly voltage/current combination.
Oh, so it needs a 5A power supply. The problem is that usually requires a >60W supply, which will be willing to provide 5A. It also requires a marked cable. I think it can legally be done with PPS if power supply supports 5A.
It is weird that Pi5 doesn't do USB-PD 27W supply which is common. It would require converter from 9V. But that would cost money.
PPS is getting more common. Quite a few phones need PPS for fastest charging so seeing more chargers that support it. It is still cheaper to buy official charger but barely. And there are multi-port chargers where might be able to support multiple Pi5.
It wasn't my assumptions that was wrong, my research (early days of the Pi5) suggested it. Which matches the github issue where the belief was that PPS was supported.
But even if that was the case you shouldn't blindly buy one where the specs match.
Is it good quality, is it made for 24/7 operation, does it get hot etc.
In the Pi case it is quite simple. The official one isn't expensive and it is ubiquitous. It runs fine on most supplies though if you need one in a pinch. But I do recommend going for the official one for long term use.
The worst part of a Pi is the SD card. It's truly the worst interface to use for booting off. Extremely unreliable and due to kernel bugs, having your system on an SD card is extremely unstable. (but that is more of a Linux issue than an Pi issue).
I found RPi to be unusable for anything remotely serious. Recently I got RPi 5 with NVMe hat. Had a lot of "fun" finding drive that will actually work with it, so more money and time spent. Got it working and found that it randomly dies after couple of days. The time and money spent I probably could be better off just getting one of those N100 mini PCs.
For dabbling with electronics Pico or STM32 seem more reasonable. RPi GPIO is too limited for anything that could use its processing power. Not sure if there are even any distributions that would support realtime operations with those pins or things like DMA, custom protocols working at 1x-1xx MHz speeds.
You can get new Mini PCs for not much more than a Pi 5, especially if you want an 8GB model, case etc. $150-$200 will get you an okay bottom barrel Intel PC with support for stuff like SATA and M.2 SSDs which are more annoying to have on the Pi.
And $150 is overstating it. Look up ‘n3350 all in one’. Lots of models and retailers selling these atm. Intel must be selling these cpus for a few cents given the price for these complete systems with storage ram, case and power supply is ~$65 even when not on sale.
Not exactly salvage, but more like buying in volume from government entities trying to liquidate either actual surplus or fairly new but legally required retired machines. The feds go through the GSA, state and local do their own thing for the most part. It's not unusual to see agencies attempt to sell multiple palettes of computers or just about anything you can imagine for next to nothing. Example: https://www.gsaauctions.gov/auctions/preview/288722
I personally disliked dealing with federal agencies, but entities as local as a school district can easily have a palette of relatively new machines, sans hard drive, and are far more flexible in payment and how you choose to get it to you. For a brief period I used to rent a truck and do weekly runs from Brooklyn mostly to Maryland and Virginia and Pennsylvania and get back to my one bedroom with 90-120 computers. At the palette level prices can get down to the $10-$20 each machine level (probably higher now due to inflation), hard drives aren't really that expensive either. The biggest headache was shipping but the process is likely far more streamlined today. Of course, it probably would've been even cheaper if I didn't operate out of a tiny 1 BR in Brooklyn and can actually own a vehicle and have reliable parking It's not exactly my tax dollars at work, but effectively it is a sort of subsidized sale at the taxpayer's expense, Intel isn't really selling anything for pennies on the dollar, but pretty much every municipality and county government will, at least at some point.
I honestly don’t get it at all considering the sheer volume of these on the market right now. Is there any chance intel would run off an old low cost design that doesn’t compete with the high end just to keep the 14nm fabs busy? It seems like there’s too many on the market for salvage alone to explain it.
The mini-pc market for those HP machines is fairly in demand and you can get them very well priced.
They're fantastic as a low cost machine with decent specs that take up minimal room, they're also usually corporate devices in which there are plenty of drivers that work well out of the box.
I bought one to act as a home server and it works flawlessly running Proxmox, over the machine I built many years back which was slower and bigger.
I just think comparing the Raspberry Pi to these Mini-PC's is just a silly comparison to begin with.
I don't see what's wrong with old second hand power supplies. These corporate mini PC's are manufactured in such large quantities, that finding them in good condition from reliable sellers isn't exactly hard.
For most self-hosting purposes mini-PC's are often the best option in my opinion, because they're small enough in size and low in power consumption to be comparable to raspi, but powerful enough for stuff like running a virtualization environment, and support containers that are only for x86. Admittedly I did go for desktop though, because I wanted more storage, dedicated GPU and ECC RAM, but for lots of people a single mini-PC could host everything.
I still like raspberry pi for doing small things where I want redundancy, like running DNS server that stays up even when I do maintenance on my main server. And obviously it's great for embedded projects and such.
Everytime I used a pi the cabling ended up being a mess. A regular computer usually have all ports on one side only, and possibly additionnal usb / trs on the other side. On the PI power/hdmi/TRS is on one side, usb/ethernet on another, GPIO on another and more easily accessible from the top. This is my major PITA concerning the PIs.
In the end most of the gains made from the small footprint are lost because the actual physical footprint increase from the number of cables you have to attach to it from all sides.
> Using an old second hand power supply is enough to turn me away.
> And the PI has many advantages. Power supply dies? I can order a new one in literally seconds'.
This is a quite a stupid argument.
a) it's totally the same for any other PC: you just order another 'in literally seconds'
b) if you don't like a second hand PSU then order a new one in the first place
> And meanwhile I wait for it to arrive I can use a spare notebook charger or whatever
... just like you can have a compatible charger for a mini-PC ?[0]
> For a home server I would recommend something beefier, an old desktop will be superior to any mini-pc and the Pi
For most of the people there is no need in 'beefier', 32Gb RAM, 256-1024Gb SATA/NVMe is all they need.
> The point of a Pi for me is more that you can have it where you need it. Attached to your TV or whatever.
Anyone can have mini-PC where they need it. It's because they are mini, not a desktop ones.
[0] by the way, most of the time those mini-PCs have a notebook style external PSUs (not some anemic square brick of 15W) and they are quite rare to break
The argument was that you already have a PSU at home. Or in my small town I can get one in 15 minutes, or order it and have it delivered tomorrow.
On the contrary, the notebook style external PSUs are more likely to break and much harder to find replacements to. When it has happened to me the best source to buy has been to order it from another country.
If you’re in a small apartment or sensitive to noise, I think it’s still worth considering the Pi or a new fanless N100 system.
I bought a tiny Lenovo i5-6500 system on eBay, and while it’s fantastic from a price/performance perspective, you can still hear a subtle whine when the ambient noise drops.
Which makes total sense - the acoustic output is probably not a major consideration when they’re optimizing for footprint and cost.
> I think it’s still worth considering the Pi or a new fanless N100 system. I bought a tiny Lenovo i5-6500 system on eBay, and while it’s fantastic from a price/performance perspective, you can still hear a subtle whine when the ambient noise drops.
This. I basically made nearly the same comment somewhere else in this thread. But then we have similar nicknames so...
Buyer beware - Dell mini PCs also have this problem, they run constantly and are clearly audible in a quiet room. The BIOS doesn't have any options to disable or reduce the RPM, software sensors don't see the fan and if you unplug it or try to undervolt it the motherboard panics and doesn't boot anymore. Had to sell mine away.
Otherwise they're a nice piece of kit. Perhaps someone can hack the BIOS to remove the fan protections.
This is situation I am in and I went for a completely fanless mini ITX system that works very well for me for close to 10 years already (yes, it is time for an upgrade - probably to a fanless N100 based mini ITX). These systems are a great alternative to both Raspberry Pi’s (that now need fans) and those repurposed office PCs, even if you do not mind the fans.
It’s an i5-6500T, which is the same as the regular Skylake, but limited to a max of 2.5GHz
I actually solved the problem by just discretely routing an Ethernet cable to a closet - but considering that it averages 10W from the wall, I definitely suspect I could get away with taking the lid off and adding a larger heat sink.
If you ever plan to use any of these older PCs: Disable Spectre and Meltdown mitigations! As a bonus you should also remove CPU microcodes from the BIOS/UEFI and make sure that no microcode is loaded via software.
The performance gain is huge if done correctly. There is no need for these mitigations on homelabs.
I find TFA interesting because I've got at home, both on my desk and in the closet, a mix of Raspberry Pis and... HP EliteDesk mini PCs / NUCs. I mean: literally the same HP EliteDesk as pictured in TFA. I bought three HP EliteDesk that were decommissioned from a nearby NATO base (so I bought them factory reset and without any hard disk in them).
One advantage the EliteDesk do have is that they are not ARM, which can help at times when I need to run that's only shipped as an image and that wasn't compiled for ARM. I know there are other ways but, well, when that happens I just run that on the EliteDesk.
Now the very obvious advantage the Raspberry Pi have: no fans. That is, to me, a big one. A huge one. My main PC is so quiet that I do hear the EliteDesk's fans when I turn one on. Typically I'll just run a Plex server at home on the EliteDesk and only turn it on when I want to stream something: the rest of the time they're off.
The Pi do run servers not requiring lots of CPU: like an unbound DNS server.
I don't see these as mutually exclusive: they can be complementary.
If I had to pick only one I'd probably still pick a Pi though.
The Pi 5 needs a fan, unless you can find a special case that allows the device to run without a fan. I run the Pi4 which hosts this blog fan-less and that works fine, even 'right now'.
Not the parent, but I run my Pi5 without a case, out of sight. I do have the Active Cooler fan on it, but on normal low intensity usage the fan turns on only during bootup and otherwise stays off.
New vs used is mentioned, but kinda critical for the comparison. Yes a $60 board is definitely less capable than a $300 PC. The cost difference is primarily driven by the factors compared: better CPU, better IO, and more memory. You get what you pay for and you can get deals in the used market.
Pi's are great for their ecosystem, being fanless, and cost for a brand new device.
Aside: ESP32/ESP8266 have taken over a lot of the hobbyist realm for connectivity + GPIO. $3 dev boards that are plenty fast for almost any single use-case scenario.
In the article, the author found a mini-PC at a lower cost than the Pi, when accounting for all the accessories he would have needed.
I gave up on Pis when I could not longer get them easily in my part of the world. And I agree with you, ESP32 boards are perfect for simple GPIO+Wifi connectivity.
I don't have room for that tinyminimicro pc case. I could fit 8 pi's in it.
> The Pi 5 can be fitted with an NVME SSD, but for me it's too little, too late. Because I feel there is a type of computer on the market, that is much more compelling than the Pi.
My pi4 has been running from an ssd for years now, no sd card. Usb3,not nvme but still good enough for my (most ?) use case.
Agreed. The whole premise of this article is absurd. An apples to oranges a comparison. The Pi is a platform for embedded systems development and design. And is excellent for what it is designed for. It's not a desktop workstation.
These conversations around the rPi make no sense to me. To me the value of rPi has always been:
* Simple to install a well supported OS with a full UI.
* GPIO array (including pwm, i2c, etc)
I could take one. Connect a keyboard and monitor. Jumper a few pins into a breadboard with an LED matrix, and right a python script to bitbang a multiplexed LED array
The only other product I have seen that I view as even competing in the same space is the OrangePi line. However, those are vastly inferior in terms of support.
Compared to an x86 PC which is what TFA is showing ? The only argument in favour of the RPi is GPIOs, but "simple to install" and "support" definitely are a thousand times better on standard x86 PCs than on any of these ARM SBCs.
It is odd to me that people call the RPi 3/4/5 "not capable."
It depends on your needs. I find it perfectly capable when I need more than an STM32/ESP, but less than a PC, which is about 90% of the gadgetry I find enjoyable.
I don't understand all the anger. If you need a different platform, you need a different platform. /shrugs/
I'm starting to wonder if we're seeing the ramping up of an Arm vs. Intel war at the same level as Mac vs. Windows, or emacs vs. vi, with Raspberry Pi as a convenient proxy, and all the usual foibles of every other us vs. them war.
It wouldn't surprise me if paid influencers were fanning the flames, either; certain folks have a lot to lose if arm goes mainstream on the desktop.
And then there's the recent Raspberry P-IPO or whatever it was, which seems to have pissed a bunch of people off...
The timing is interesting.
I can see the appeal of the N100 and similar -- and am currently considering one -- but it's hard to beat the ~1W DNS/DHCP/Consul servers that have been humming away in this house trouble free for the last decade...
I don't need those to power my homelab. There's a reason they're separate, and tiny low-power Pi's are ideal for that job.
I have in my possession more than merely a hammer. :)
Where are these tiny low-power RPis that idle at 1W ? Are you measuring at the wall (i.e. with the power adapter inefficiencies?)
The entire point of the article is that at least the non-micro RPis are not "low-power" at all (something that agrees with my observations). The article is quoting around 3.5W for the RPi5. I was also getting a similar reading for my older RPi4 after months of fine-tuning, while _out of the box_ an N4000 miniPC had 1.8W consumption at idle. RPi may be cheap, but not much else.
I'm not running 4's or 5's. I'm running 2's and 3's. And they're mostly idle; DNS, DHCP, and Consul are all extremely lightweight. No video, wifi, or anything else; just ethernet and an SD card.
No sense in running anything more powerful than that. Having the latest and greatest -- and top-tier performance -- isn't always necessary.
And I still consider to be a credit card sized server to be tiny. Though watch-sized machines do take the cake there...
The Raspberry Pi 5's official heatsink comes with a fan and its collection of software is dwarfed by what's available for a x86 PC regardless of whether it's running Linux or Windows.
I've been running a fanless mini-pc as a firewall for years. They work just fine. The ones you get from aliexpress come in a case where half of the case is a massive heat sink, and it will get a little bit warm.
Pi will automatically throttle before running too hot. And in any case, I doubt most people run their Pis hard at all. I bet most use cases will be completely fine even without that heatsink, with no compromise on the lifespan.
I've got one of those, and it houses a system with 8 CPU cores, 32 GB RAM (can be upgraded to 64 if need be), 1 TB NVMe and 4 TB SSD - and it's all inside, whereas with an RPi the SSD would have to be external. The only thing that's collecting dust now is the old RPi.
It provides a WebRTC stream for a USB camera (or Pi Camera, what I'm using). Rather than the old, inefficient, low-quality MJPEG stream. The software itself will run on anything, but the WebRTC only works on a Pi for now.
I always keep coming back to RPi because of the software support. Now when I think about it, basically everything is secondary to that. Most of the stuff with RPi just works like charm, and when not, you usually find tons of information how to make it working. Not so with other devices
The article only compares idle power as far as power consumption goes, but regarding compute performance compares the systems under full load. This is somewhat of an apples-and-oranges scenario. If you care about the compute performance under full load, you should think about performance-per-watt, not about idle consumption. And in that capacity, the RPi 5 will easily beat any enterprise surplus mini PC with a 10-year-old architecture.
One of the mini PCs tested in the article has a CPU with a 65W TDP! That's considerably more than the Pi 5 will draw (the entire SBC's power supply only has a capacity of 25W).
Pi's are great for easy hardware hacking, but I don't know if they ever made that much sense as home servers. You could always pick up used office/minipcs for even cheaper than a bare pi board, and if you picked carefully, you wouldn't really be using much more idle power.
Also €100-150 for those used 1L boxes sounds a bit pricey to me, since in that range you can buy brand new minipcs that perform similarly (personally for a network-centric device probably I'd go on aliexpress and grab one of the fanless N100 router-style pcs).
They are ridiculously overpowered for a number of usecases, even the older and cheaper 3 models. I'm running progscrape.com on a 4 and it held up to HN traffic without sweating at all.
I had a Pi1 running Stylus for home monitoring with maybe 10% CPU use at any time.
I've moved my media players and servers to mini PCs and unlocked Chromeboxes and couldn't be more happy. Performance is on another planet compared to the RPi.
Mini PCs however can't be used when one needs a small board with lots of gpios, but there's a lot of (often cheaper and faster) competition in that field as well.
Please, keep in mind that the usual story "other boards don't have decent Linux support/community" is simply not true. Here are Armbian and DietPi pages where you can find images with mainline support for a lot of common boards, including forums. Images supplied by the board manufacturer should rather be intended for quick testing only as you can't count on their support; just ignore them and go straight to Armbian or DietPi sites, and consider contributing for their hard work.
Powertop is a great utility for identifying wakeups and CPU C states, but for tweaking power management flags in Linux, I find TLP (https://linrunner.de/tlp/index.html) to achieve greater power savings at the "cost" of more configuration.
It wouldnt touch a data centre but the simple ability to run CUDA (Rapids.ai for example) would mean you could prototype things pretty efficiently all on a local setup and not have to pay for GPU costs.
In keeping with the spirit of the Pi that sparked this discussion (I mean, a little bit), an NVIDIA Jetson AGX Orin Developer Kit with 64GB of RAM is $2000, and the Jetson Orin Nano kit is $500.
Yeah thats ridiculously expensive for a home lab IMO, the thinkcenters can be picked up for a couple of hundred bucks used and a T400 runs about the same. So $400 for an AI capable home bench would run $1200 for a 3 node cluster - I can live with that.
My 5 node Rpi5 Cluster ran $1500 with NVMe Hats and a PoE hat and cant do any GPU work :(
The AGX Orin has over five times as many CUDA cores as a t400, and the dev kit is designed to do pretty much exactly what it sounded like you wanted. The other path does offer the opportunity to wire together a pile of new and used equipment and generate heat with it, so if that’s your idea of fun, I guess the choice is clear.
Used, many years old PCs are cheap compared to new, non-used ones and can be still perfectly sufficient for basic tasks. True, but also seems very apples to oranges.
The proper comparison would be either be pi vs new mini PC (e.g. n100 based) or new mini PC vs used mini PC.
I recently spent some time looking into exactly this. I ultimately decided to go with an N100 running Proxmox. It is a wonderful compromise between power utilization and compute.
The N100 was roughly $50 more than the Pi 5 (after adding storage to the Pi). The idle cost to power both is roughly $5/year for the Pi5 and $10/year for an N100 (based on local electricity rates YMMV).
I'd argue the Pi still has a purpose: Running it off of battery/solar is likely better, it is physically smaller, and it also has pin-outs and documentation/software to support it. If all you're after though is a small-low power usage computer, it may not be the first choice anymore.
Raspberry Pi is mostly an expensive toy at this point, it costs 60-80$ and it's pretty hard to buy one to begin with. You can't use it for most electronics projects because it's overkill and also it's not as powerful as a n100 or a second hand pc. If you are lucky, you can snipe a bunch of ryzen 3 hps or dells (if you don't need a lot of performance) for 60-80$ or ryzen 5s at 120$ (if you need a bit more).
This headline makes as much sense as "the Kia Niro is no match for the Volvo FH" -- it really all depends on what you want to do with it?
RPi5 is a great platform for prototyping, and many hobbyist applications, even if you move to ESP32 or similar afterwards, or if you decide that PC-ish platforms work better for you after all.
"One size fits all" has never worked in computing history and most likely never will...
In my area there are a few pawn shops who offer second hand DELL enterprise pcs, which is great for me because they struggle to sell them to the general public and I've gotten my hand on one or two for a steal.
I use one as a server, and another as a media center PC. My server has been going without issue for 5+ years now (granted, I'm not stressing it, its just for serving files and doing downloads).
Not that different for the prodesk/elitedesk small form factors IMHO. There’s a steady supply of cheap used ones available on most online marketplaces (there’s probably millions of them being cycled out of offices every year).
Get a new replacement one, swap over the NVME. Doesn’t have to be the same cpu, linux handles the rest.
I'd wager that most people are not using any GPIO on theirs. Typical usage is likely a low power computer they can optionally plug a USB cable into; 3d print controller, home assistant w/dongle, etc.
I'm using AVR/STM for IO. have plenty of them around. Just loaded with a basic serial to IO passthrough program.
Can't complain for £2 each on aliexpress.
But what do you use Pi GPIO for? Using Pi GPIO directly leaves pins unconfigured or stuck while your app is inactive. Aren't most need for GPIO better served by Arduino + PC?
I have a Pi Zero W that is a spooky-good GPS-backed NTP server. It relies on GPIO for getting tightly-accurate PPS pulses from the GPS module (which a USB-connected Arduino won't help me with -- the timing would be much sloppier).
I have also used a different Pi Zero W with an SDR dingle to decode APRS weather data, while also using GPIO to read a local DS18B20 temperature sensor and to switch a solid-state relay.
I could have done this last thing with any random Linux-ey PC plus an Arduino, but then I'd have two problems. (It would have also cost me rather substantially more money: I bought these Zero Ws very, very cheaply from Microcenter.)
Cute, but missing some things that real ntpd on *nix offers.
And I like ntpd. I also like self-contained systems that I can tweak without using a compiler and a dev environment. I'm a fairly competent computer user, but I have zero aspirations of being a programmer when I grow up.
(If a Pi Zero W can do the job, then: Why must it not do that job?)
It's all good. Lightness certainly does have advantages in many circumstances -- I've got some projects in the works that use ESP32s and Pi Picos. But once I set them up, I never want to touch the code again; adjustments will be made in "userland" using a web UI. (Remember, I don't want to be a programmer when I grow up.)
Thanks to these discussions here in these threads, I've upgraded my Raspberry Pi 4 router-box (which has cheerfully been running OpenWRT for years) to the latest stable OpenWRT. It took some time to get this done but it is done, and I managed to get it done without tossing my existing configuration or disabling the household's Internet access for more than a few tens of seconds at a time.
And I've also gone through the NTP config on that router, which [previously] was never quite right -- mostly, due to complications with BusyBox also knowing how to be ntpd (ffs) and poor system documentation (which I think I may spend some time improving).
Sure, there's other ways to do all of these things... but Raspberry Pis are fun for me in ways that other hardware (black-box streaming devices, black-box routers, even black-box GPS clocks) lack. I even take one or two of them car camping with me, where their somewhat unique ease of downloading and flashing an entire pre-configured specialized system image onto a different easily-swapped, inexpensive SD card can be tremendously useful[0].
[0]: Stuff I can't do with an MCU, or with a regular PC. "Oh, this carefully-prepped LibreELEC Kodi can't work for Movie Night in the Woods because the USB stick I carefully copied a bunch of movies to is somehow empty and the data that should be there is nearly three hours away? Let's just take that card out (preserving that configuration completely), and follow the directions to get a decent Plex client working the Right Way on a different cheap-like-chips SD card. I can do all of this from my phone."
I don't think you understand how GPIO is used on a Raspbery Pi. You can flip GPIOs in a BASH script, or set events on them in Python. GPIOs are first-class citizens in Rpi land, which is what makes it so convenient: you get GPIOs + Linux, seamlessly.
I have a Pi4 hooked up to a motion detector, a camera, a solenoid and some SCRs for a floodlight. A python script waits for an event interrupt from the motion detector on a GPIO. Then it turns on the lights and opens a door via GPIO, and starts recording in another thread (via a pipe to gstreamer). When the motion times out, it pushes the video to an private S3 bucket (using my own handshake for a token so that it can't be spammed).
(I used to have some GPIOs connected to motorized cat toys so that I could trigger them remotely while watching an realtime stream, but that throttled the above camera which was more important.)
... It is also a BLE gateway that monitors 4 temperature sensors, and also pushes that data up to the cloud with another python script.
... and it also is an MQTT gateway for a few NXP WiFi devices scattered around the yard, and their data ... you guessed it... is pushed up to the cloud (they control sprinklers because BLE doesn't have the range). I can log into my personal website and turn sprinklers on from anywhere my phone works.
Sure I could do the WiFi and BLE with a PC, but why waste the energy and space when I can use a tiny Pi that is already doing a bunch of other things?
I just keep adding things to it because I have my own cloud system (HomeAssistant is a big fat bloated joke).
Having a good BLE sensitivity and WiFi chip on a tiny linux board that runs dozens of python processes listening to GPIOs, and running Gstreamer would be too much of a hassle with a PC and far beyond an arduino.
I buy every Pi and try to run MechWarrior 2 in dosbox and they've all fallen short. I also have a Surface Go 2 (bought used for $80) w/ 4GB of memory + Pentium Gold and it runs flawlessly.
Relatedly- I find that old hardware is still very capable but ruined by new media codecs that lack HW acceleration. Are the new codecs really worth it when they essentially render whole generations obsolete for media?
I also have recently gotten an N100 mini-PC, which helped me put one of my ancient mini-ITX i7 machines out to pasture. Those are amazing (and are dipping under $120 with 12GB RAM and 512GB SATA SSDs).
But more to the point, I think the space the original Pi occupied in hobbyist land is being eaten up by RP2040 and ESP32 MCUs, which can do everything I need I/O wise and with increasingly sophisticated software support (MicroPython is amazing for prototyping, and I even got a WaveShare RP2040 board to test that is a direct drop-in replacement for a Pi Zero)
I ended up swapping out my retro pi for one of these micro PCs. I’m mildly sad that it’s not the same form factor, but once I got over that, I became a lot happier. The Pi (3) didn’t behave as smoothly as the Lenovo I replaced it with.
I love the Pi and the flexibility and community it has. The only thing I’ve seen to rival it is the Arduino. But for casual needs like home theater or home assistant, these 2nd hand micro-PCs are excellent.
Quick aside, I actually got the micro-PC for a Starship bridge sim setup. I never put that together, but it is also a great use of these.
I've gone this direction with my home lab stuff as well. I have an assortment of older RPis up to a couple RPi 4Bs. Most I bought to use on projects where I wanted Linux plus GPIOs and they were and still are good for that purpose.
I've also given in to the temptation to use them as little servers on my LAN for various sorts of things. I've even tried to use them for HTPC boxes and they have worked ok.
However I have sent all my Pi's back to projects where I just want Linux plus GPIO. For everything else I replaced them with cheap mini PCs.
1) a bare Pi on a workbench/desk is fine but when connected to a TV in the living room has a very low WAF.
2) the arrangement of ports on a Pi are a complete pain in the ass. Even with a case any installation of a Pi looks like a careless hack job.
3) MicroUSB is complete fucking garbage. The port/connector wears or breaks and the slightest bump powers off the Pi. MicroUSB power supplies are usually also garbage and typically have criminally short cables. "Appliance" appropriate power for a Pi costs more than the Pi itself. USB-C in the 4B is better but I've only got one of those.
4) like power supplies a decent microSD card that won't randomly fail in a Pi ends up costing as much as a Pi.
I replaced a bunch of individual Pis running servers with a single mini PC with an N95. The Pis weren't taxed with the server loads so the N95 doesn't break a sweat. It's also way easier to manage since the various server apps are managed by a docker compose file.
The mini PCs replacing the HTPC Pis are just a better experience overall. The ports are all on one side, they have same barrel connectors for power, and their local storage doesn't magically corrupt itself because the unknowingly cheap power supply didn't provide enough power.
I still love the Pis for small nerd projects but just don't want to deal with them anymore for pretty much anything else. If you don't need GPIO for a project and physical volume isn't a prime concern a <$100 mini PC is a far more convenient option today than a Pi. That wasn't true ten years ago which was why I started using Pis for servers and stuff but today it is definitely the case.
Originally the server & VMs were running on my windows 11 pro dev machine. Keeping HyperV made it easier to move them to another machine.
Now that I have a separate server, I find it easier to develop the VMs on my dev machine, snapshot and then migrate them to the server. I have a library of base images for alpine, Debian & Kali Linux that I can launch as easily as AWS.
Using Hyper-V across both makes this seamless. I can manage the servers using the Windows Management Instrumentation (gui) or RDP into the server.
I know many Linux users have aversion to Windows, but I get a lot of benefit out of Windows for other applications as well (gaming, one drive, document indexing, copilot, as well as all the great native apps)
Has the Raspberry Pi tripled in price since launch? I was skeptical of all these threads of “this is better” but it doesn’t seem like it’s as ridiculously affordable as it once was.
I feel these comparisons are a bit funny. The "RasPi 5 is no match for X", showing a picture of a computer that seems around 5-8x times bigger in area, and probably around 20x bigger in volume. They also cost A LOT more, but if you get them second hand only +50%. Color me surprised, even when ignoring the general totally different markets of those, even as a tiny PC these are not fair comparisons (specially comparing a second-hand device to the full cost of a new Raspi, adding the accessories that we all hackers/makers already have).
Yes, for ALL purposes that people use the Pis for there's definitely a better, more specific option. But that's the great thing of the Pis, that they are very general-purpose for hobbyists, while also being very well standardized and documented (and run Ubuntu).
I will always love raspberry pi's, they are well supported and great for quick and cheap prototyping. I have used them to teach children how to code (They really enjoyed working with sensehat), research at university, and at work for rapidly prototyping. The size is also easy to work with.
There are better SBC's out there, but raspberry pi is familiar.
I really hope the recent IPO doesn't change too much...
Has anyone else had thermal management problems with the pi5? I have one running just a couple of servers tucked away on a shelf and a corner of a room, and I find myself needing to physically recycle power on the thing about once every few days (This on top of the daily auto reboot script I have set up as a cron job). I suspect the Wi-Fi is overheating it.
Also check your USB cables if they're supplying power. Low-quality cables are notorious causes of instability with older Pi's, and I doubt the 5 is any different in that respect.
I read here that people discuss the current prices of raspberry pi. The thing is... it doesn't really matter now.
People I know that wanted to play with single board computers, they bought raspberry pi. The market is saturated, and PIs do not break that often if you do not use pins.
It will take time for any other SBC to replace it, or even to come close.
This is timely, after not getting my Orange pi to boot an image, I saw I can get an actual mini PC like these used on eBay for like $60 shipped. 8GB ram and 500gb SSD with small form factor. I'm considering buying more and trying to host some old school LAN parties as used LCD monitors can also be had for about $50.
My home router runs on a dual core Intel Celeron. Software wise it’s OPNSense in KVM and a few docker containers on the host. Measured power draw is 4-5W. That’s with an SSD and 16GB RAM. It’s also way faster than any Raspberry Pi. The notion that Intel draws a lot of power is wildly outdated.
I have some of these mini PCs at home. Not just HP but Lenovo and Dell are also making them. A Dell Optiplex 3080 with i5-10500T, 16GB DDR4, and 256GB NVMe SSD is ~200€ and can be upgraded to 64GB RAM with a lot of storage (NVMe + a 2,5")
Pi 5 is also no match against your old iPhone you no longer use, and you would get approx the same credit value towards a new phone (as the cost of a new Pi).
for using it as a server for self-hosted services, sure. but using it as an embedded device, almost impossible owing to the lack of IO pins.
people that use the raspberry pi to host their things should definitely use these mini computers.
i believe there are other great alternatives out there that's just as affordable (cheaper than a Pi), just as powerful as a mini pc, and has an extensive IO support for prototyping.
I wrote a post on my (deleted) blog back in 2022, and in it I documented some of the issues I ran into with the Raspberry pi 4 when I was building a media pc that I bolted onto the back of a TV.
Briefly, the Pi absolutely sucked in a bunch of ways compared to the MeLE Quieter 3Q that replaced it.
1. Pi couldn't use USB 7.1 audio device in the mode I needed (7.1 out + line in). No idea why, but Linux totally froze up when I selected that mode, and remained frozen until I physically unplugged the usb device. On the same USB device this worked fine on other computers. This meant that I couldn't use my external Bluetooth receiver seamlessly via the Pi.
2. YouTube failed in a bunch of ways. Seriously! Neither Chrome nor Firefox would work 100% reliably. Sound issues on most videos, awful performance, locking up the browser.
3. Can't be put to sleep or remotely started using Wake On LAN
4. Lacking sufficient hardware encoders/decoders
5. Everything goes through the USB bus, so you get awful storage/network performance
6. Can't use Wine to run windows apps. While this may not be an issue for many people it was for me because I wanted to use MusicBee. Wine works fine on the Pi if you've got ARM compatible windows apps. So good luck with that.
7. The hassle and cost of getting a decent case.
I ended up spending a few hundred bucks on the MeLe and everything just worked. Flawlessly. First time. And I had a wide-open choice of Linux distributions.
Well... Yeah... I have a bunch of raspberry pis from back when they were cheap + a raspberry pi 400 I won at a hackathon a few years ago. There are things about them that I love(the older ones to be more specific): low power and great thermals. But ever since their prices skyrocketed, I have completely abandoned the newer ones and opted out for what me and my friends refer to as "cubes". Optiplex micros and similar ones. They are a tiny bit more expensive but you have expandable storage and memory and in lots of cases upgradable cpu if it comes down to it. And most importantly x86, which I take over ARM any day of the week, especially when I have to dive a layer or two deeper - it's just more convenient having all the additional tools.
However I have one or two projects that I keep kicking down the road due to time constraints which involves a rasprerry pi zero w2. The reason being is that it's compact, packs a decent punch and is extremely easy to power it from an 18650 battery and keep a low profile.
I'm fine with the current prices of the raspberry pis but it's hard to justify using one of them if you don't need GPIO pins or something that uses little power. My advice to the raspberry pi foundation is to stop trying to push performance and focus on features. I'd be the first in line to get a new raspberry if it comes with a battery hat and LoRa built in(the project I was referring to really). Here's another one(which ironically already sort of exists): the milkv duo. There are lots of cases when you don't need a full operating system 24/7, which would waste a lot of resources to do one-off tasks every now and then. Ideally you should be able to boot the OS from a sensor or some sort of signal, which is hooked to a very under-powered micro-controller only when you really need it, let it do it's job and then power it off. As we stand, the raspberry pi foundation largely offers the same product in different form factors with some arguably marginal upgrades, which most people don't really need.
1957 single core passmark and 5548 multicore. More importantly its 5 watts at full CPU
You can run it off 19v DC, and its physically tiny. Yes, it has a fan, but its not that loud, significantly quieter than the HP
The second hand HP is just not compelling anymore, unless you need more than one drive, or you need something physically larger.
TLDR: the PI isn't great for KVM. The n100 is.
I have a mix of n100s, NUCs and Pis. Each has their purpose. If you really want cheap ephemeral linux, then the pi-zero is still dirt cheap. £20 for a linux machine with wifi.
Depending on what you want to do, the Pi may still be a good option, yet I think it's good that people are aware of alternatives such as these second-hand mini PCs.