Who Ate All The Pi?

Wednesday was the last day of February, and leap year questions aside that date marks the anniversary of the Raspberry Pi launch. The oldest commercially available Pi is now 6 years old, and to mark the occasion the Raspberry Pi people have put up a retrospective of all their different models.

There is a primordial prototype from [Eben Upton]’s bench that involves an Atmel processor, but the first board dangled in front of the public was a Broadcom one, the BCM2763 ‘micro DB’. This was a form factor like one of those Android TV sticks, and while it was not a Raspberry Pi internal design or indeed sporting the SoC to be used by the Pi itself, it was sufficient to capture the imagination of what would become the Raspberry Pi community.

If you got out of bed early (British time) on the 29th of February 2012 and tried to order one of the first commercially available boards, you were most likely to be out of luck. The relatively small first batch from China was oversubscribed massively, both the RS and Farnell websites went down completely for most of the day. We received our model at some point in May. It’s an over-used phrase, “And the rest is history”, but it seems entirely appropriate here. The Pi has passed through several iterations and increased in both computing power and memory, it has spawned a whole industry of peripherals, a huge community, and a host of competitors. We have quite a few of the boards in the blog post, but some of the more exotic ones have evaded us.

It’s not the best or most powerful board out there, many of its competitors can beat it on performance, but it remains the one to beat in small and cheap Linux-capable single board computers. Why is this the case? It has probably the best-supported Linux distro of all of them, and that community has already answered many of the queries you might find with your board.

So there’s the story, a successful product line, community, and foundation. The Pi blog piece is very much their PR, but it doesn’t need to gild the lily. However, that will not stop competitors from taking aim at its crown, and the field remains open for one of them to topple it. Which of course makes for fascinating stories for us here at Hackaday, so we’d encourage anybody with an electronics factory in China, a bright team, and some good ideas to give it a try. Meanwhile, we’ll be looking towards Cambridge for whatever new products will sport the fruity logo.

44 thoughts on “Who Ate All The Pi?

        1. God is alive; Magic is afoot
          God is alive; Magic is afoot
          God is afoot; Magic is alive
          Alive is afoot…..
          Magic never died.
          God never sickened;
          Many poor men lied
          Many sick men lied
          Magic never weakened
          Magic never hid
          Magic always ruled
          God is afoot
          God never died…
          -Leonard Cohen and Buffy St. Marie

  1. I happen to have an original model B. :D and it runs the current System well enough. Tho I have a Model 3B now also because I need more speed.

    (Interestingly the Pi Zero uses the same CPU as the Orignal.)

    1. I have a Raspi A, 2 Raspi B’s Two Raspi B+’s, A Raspi 2, 3 Raspi 3’s, 2 Zeros’ and 3 Zero W’s …. Actually the Zero’s CPU is better than the originals as the manufacturing process got better and they run at a higher clock speed and the Zero W’s easily have enough power for OctoPi which replaced the Raspi 3’s on my 3 3D printers

      Of course I didn’t let those Raspi 3’s go to waste, one is already used in a arcade console running Retropie, a second is my current project for my boat with has Temp, humidity barometric pressure heading sensor GPS module and an AIS receiver module hooked to the Raspi 3 running OpenPlotter which outputs the NMEA data from the heading GPS and AIR receiver to my sidescan finder/chartplotter and send that data along with the weather data over WiFi to a Windows tablet and Android phone and tablet. The remaining Raspi 3 will become the main processor in an advance weather station im designing later this year which has the normal sensors plus UV, IR, CO2 and a lightning detector.

      I love these things because there is so much software support out there and being many a hardware guy that’s important to me. If you can think of an application it’s almost 100% certain you’ll find some example code you can use to develop and refine your ideas. It sure seems like it was longer than 5 1/2 -6 years ago I first got one of these to make an XBMC box for my living room ….. Although it always seemed like years between the time a new model came out and when you could finally get your hands on one

      I have to laugh every time a copycat comes out with the label “The Raspberry Pi Killer” and then doesn’t even come close in terms of support and actual usability

  2. I have one of the initial 10 000 pcs run. When I got it I booted it up, played around with it for an hour or so and then thought “meh…” and put it back in the box and chucked in it a drawer. It took me until just a few months ago to purchase another pi – this time for a digital signage project.

    1. the pi definitely needs a purpose before a purchase. if you purchase it on the official guise as an “educational computer”, or a “programmers first computer” you are going to be very disappointed. i especially don’t like the python everything bs when you got a perfectly good gcc toolchain to work with, despite terrible compile times. i think they grossly misunderstood their consumer demographic. they should have gone after arduino users wanting more beef or the maker crowds, anyone looking to do a project such as yours. i have 3 of them, one being part of a pi tablet, the other 2 collecting dust in a parts bin.

      1. I bought my first Pi2B without a solid purpose. I ended up using it to flash a bricked ip camera, to seed torrent files, to do home automation and as a portable game console (“portable” as long as you ignore the 2kg SLA battery powering it) during long trips. I see it more like a toy or a tool than as an educational computer or as something with a fixed purpose. The GUI performance of a Pi is awful anyways. Also, Pyhton is annoying.

        I’m on my 4th Pi (2B, 3, Zero, Zero W) but only 2 of them are in use. 2B is still serving torrent files and Zero W is serving two personal Telegram bots. The Zero was being used as a volume knob for my desktop, but was replaced by a digispark attiny85.

        My Pi3 is “collecting dust” too, but playing with electronics, SBCs and Arduino is my hobby. For some reason, I don’t see it as wasted money…

    2. My original model b was purchased so I could have an anyways on computer for learning Linux. And it did that admirably. It was an ssh tunnel endpoint, a Nethack server, a Kippo honeypot, and more. Usually the sd card would get corrupted after a frew months to a year and it would be reborn for somee other purpose.

  3. id argue that raspian may very well be the best linux distro ever (probably because i didn’t spend most of my time screwing around with wine trying to get my preferred windows software to work and just did linux stuff with it) and my linux fu would have most definitely been weaker had it not been for the pi. its also the only os i ever wrote a driver for (diy i2c joystick).

    1. Indeed! Things tend to work more effortless when I keep my grubby fingers from those obscure knobs. All my software lab equipment is always resetable to a known state, I have not figured out of how to do this cheap with hardware though, messing up on a budget always take too much time.

      “get better” doesn’t work for me.

  4. The Pi doesn’t fit a slot I am interested in. I want either a board with a well developed gui development program like vis studio that allows easy fast programs to be made in a real language like c, c# or java , or i want a board with minimal or no OS and very low latency IO that can do rapid calculations edging toward DSP type applications and control.

    Yeah I know that there are ways to get there but I am a lazy slob and will NOT wade through piles of special setup instructions and spend days on the net finding out why things don’t work as they should. My goal is to make things work not learn the minutiae of linux or a particular platform.

  5. Pi’s are great. I use it as a torrent server/downloader, a DLNA Server, an IoT gateway, a packet sniffer, cable tester, iperf performance tool, lan mapper, SPI, I2C programmer, video streamer, NAS, Video Grabber, PVR/NVR, and just a general super low cost PC to test any kind of software.

    The quad core models are fast enough to test most linux software, without the need for a virtual machine running on the x86.

    DOSBOX Takes care of most 80-90’s and early 2000’s software, Retropie lets me eulate a NES,SNES, DS, and almost any other early gaming system I had.

    They are cheap enough where i can dedicate one to every purpose, and all I have to do is either swap out an SDcard, or a USB stick.

    Unlike closed tablets that must be rooted, and can only run limited software, the Pis can run almost anything.

    the XU4 is my next project, it should make a great high-speed NAS. With a USB hub and a few disks, I can get a few TB of storage online for cheap, without the cost of a QNAP, and without the security vulnerabilities.

  6. My 9 year old has one set up in his room, we write programs in python to check his math homework. We have gone through a scratch programming book and a python programming book together. It has been a good purchase. I think this is the use the creators envisioned and it works just fine in that respect.

  7. “but it remains the one to beat in small and cheap Linux-capable single board computers. Why is this the case? ”

    Great PR – Just look at their Wiki, is practically an ad. Originally marketed as a $20 personal computer (not PC), with operating system, the size of a credit card. Tech blogs ate it up.

    Great timing – Similar products already existed at higher prices for engineers. The disillusioned, Y-generation programmers wanting to get into electronics but not wanting to learn about hardware was a massive untapped market.

    Also, being the first means you become the de facto standard everyone supports and therefore becomes the ‘most stable and most supported’.

    Shrewed business tactics – Using government grants and manipulating non-profit/charity status was brilliant. Eben Upton said he thinks it would have been impossible as a commercial venture. I think he’s right. No commercial venture could get the first year or two of production components for free or at million piece volumes while only building thousands.

    Not being (too) greedy – Being a ‘charity’ just means they can’t make money during the ‘exit’. All the founders are still getting paid silly large salaries considering the work they do now. Brilliant move that allowed them to not be another starving startup that fades away from lack of profits.

    Creating a loss-leader (PiZ) as a market barrier to entry was also brilliant. They just have to artificially keep production volumes low and it makes it very hard for new players to compete.

    1. Yea, when you after 2.5 years after the release still can’t purchase more than one (1) RPiZero then something fishy is going on. They can pump out millions upon millions (15-20 million) of the regular more expensive models without a problem, but as soon as it comes to the cheap Z or ZW then it’s all of a sudden next to impossible to ramp up the production even a bit.

      1. Exactly.

        Naive companies incorporate a Pizero, instead of other offerings, because of the unrealistic price.

        When it comes time to buy in volume, the distributors say you can only buy one or two “due to demand”.

        You contact the foundation and ask to buy in volume. They say we only sell to distributors because of our charity charter to education.

        You request to become a distributor and they send you shady agreement terms which basically allows them to determine the delivered volume and number of units you get to sell to each of your customers. Odd.

        You ask why I can’t get more and they say “demand is too high”. Anyone in manufacturing knows this is bull shit after 6 months.

        You contact other distributors and they all tell you the same thing. Same volumes, never increase unless sales of the other boards increase.

  8. My Pi wants more vc4 documentation releases for its birthday !
    power/clock management, hdmi bringup, usb dma engine, csi, mpeg2/mpeg4 decode hw & mpeg4 encode hw.

    and since the last mpeg2 patents just expired, an updated firmware without the need for a license key would also be nice..

  9. I have two Pis. One runs my weather station and web site. The other is on the table near where I watch tv. I use it to look up stuff while watching TV or otherwise with the screen saver, it’s a digital picture frame. The display is a reclaimed laptop LCD from an otherwise destroyed laptop.

    I’m looking for the mythical $5 PI zero to possibly do some IoT projects but haven’t been able to find any still.

  10. I still have an original model B. It was an excellent torrent client until some weird I/O buffer/delay thing started happening, so I re-purposed it as a squid proxy+cache, then got bored with that and put it in a drawer until I learned about pi-hole. It works quite well, but with only one network interface it’s subject to contention, so one of these days I’m going to get another one with 2 NICs.

    Pi-hole could be greatest disruptor to internet advertising models since adblock and noscript.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.