The business arm of Raspberry Pi is preparing to make an initial public offering (IPO) in London. CEO Eben Upton tells Ars that should the IPO happen, it will let Raspberry Pi’s not-for-profit side expand by “at least a factor of 2X.” And while it’s “an understandable thing” that Raspberry Pi enthusiasts could be concerned, “while I’m involved in running the thing, I don’t expect people to see any change in how we do things.”
↫ Kevin Purdy at Ars Technica
Expect changes in how they do things.
The Pi has already stopped being a Pi.
It went from cheap £30 board to £75
My guess is they’ll now start building bigger and bigger models (inflating the prices) and probably enter the low end laptop market.
They never seem able to keep with any kind of demand. That’s ultimately going to be their downfall
$35 is probably no longer viable. There is inflation. Even if you take Rpi 4’s introduction, that $35 is $41 in today’s numbers.
That being said, yes, the no longer cater to “hobby” market. At least not to hobby market alone. Especially during a certain period, they seemed to have allocated boards for more enterprise customers, and there is a large demand for “compute models” which can be used by device manufacturers to put a brain on whatever they are building.
There is of course RPI “Zero”, which is probably the most similar one to the original purpose. It has a really low power CPU, basic I/O (no more wifi/bluetooth (at least in base model), and limited GPIO pins.
To compare, the original RPi had:
700 Mhz 32-bit ARM, 256/512MB RAM, Ethernet, adn 26 GPIO pins, two(?) USB ports
And modern Pi Zero 2 W has:
1 Ghz Quad Core 64bit ARM, 512GB RAM, WiFi, single micro-USB port
And if you want to go more lightweight RPi Pico has:
133 Mhz dual core ARM, 264K SRAM, 30 GPIO pins, single micro-USB port (SparkFun and others sell type-C versions)
So, in that sense it actually got cheaper ($35 vs $15) and slightly faster, but lost a bit of capability (no Ethernet, USB hub, or camera/display outputs on those lighter models).
(I really need that edit button)
sukru,
That’s great on paper, but I had to pay ~$80 for a RPI4 a couple months ago. This included a case and power supply, but still…double the MSRP.
The specs are fine, or would be except I couldn’t get an RPI zero w2 at all when I needed to. I ended up buying an old RPI zero w1 even though it didn’t have enough cores for my application, and it broke my heart to buy a micro-usb peripheral from yesteryear. In theory, I actually like what they’re doing a lot, but I haven’t been able to get what I need with any reliability without going to scalpers 🙁
Well, it didn’t get cheaper, they entered a new market with a new product line that competes with microcontrollers. rather than the linux SBCs they offer.
I could recommend CanaKit:
https://www.canakit.com/raspberry-pi-zero-2-w.html
They used to have a very strict “1 per order” policy in zero, in any kind. (i.e.: they didn’t even allow zero 2, and zero in the same order). But I can now add more than 10 per each.
They even have PI 5 on backorder at retail price ($60/$80). But mine has been waiting for ~3 months, and have not arrived yet,
sukru,
Thanks for the link.
Pi locator shows them instock at a few other places now too. (why isn’t canakit listed?).
https://rpilocator.com/?country=US&cat=PIZERO2
I don’t know if I should be buying more now in case they go out again.
I registered once at canakit a few years ago but I backed out when I saw their shipping prices. $15 is rather high for small orders, 100% price of a w2 (especially when you could only buy 1). To be fair, I might tolerate that given no other choice, but now that it’s in stock by multiple sellers, I can vote with my feet for a seller with better shipping costs 🙂
Alfman,
Probably because they are in Canada. That also affects shipping prices.
sukru,
I didn’t know that, but makes sense. I guess that’s how they got the name canakit.
I
I had a RPi 5 on backorder from SparkFun since they opened up orders back in November, and the ship date kept slipping further and further even though they “shipped orders in the order they were received”; if that was true I would have had mine in December or early January at the worst.
So instead, since I was wanting a low power OpenBSD device and I didn’t care about GPIO and all that, I canceled the Pi order and bought a Mele Quieter 4C from AliExpress that arrived in about two weeks from China to the US, for not much more than the price of the Pi and requisite power supply and cooling module (using a coupon that has since expired, but it may be available elsewhere for the ~$160 I paid). It is fanless and still nearly twice as powerful as the Pi 5 on paper, while only using a few more watts at idle.
Apart from a couple of GPU issues — no video using a DisplayPort to HDMI cable, but a USB-C-DP to DisplayPort cable works great, and not-great-not-terrible video playback due to OpenBSD being behind on Intel DRM drivers — it’s been rock solid running OpenBSD 7.4. Curiously I couldn’t get -current to boot on the device, so hopefully that issue is worked out by the time 7.5 goes live, but otherwise it has been an amazing device for the price and size.
I also briefly used it as a HTPC/retrogaming device, but found it would throttle under heavy load when transcoding media or playing anything more power-hungry than PSX games. It would be fine for 32-bit retrogaming and under, and playing back media directly without transcoding on the fly.
Here’s a link to a release announcement of the device for anyone interested:
https://liliputing.com/mele-quieter4c-is-a-pocket-sized-fanless-pc-with-intel-n100/
Morgan,
You are right, the “mini pc” is a viable alternative if you are looking for a… mini pc. They are more capable than RPi in some areas (like type-C video out or m.2 nvme in higher end models), and can use a full desktop operating system.
sukru,
I’ve done the same thing as Morgan for routers, not because the x86 specs were better or more appropriate, but because I had much more confidence that I could support them indefinitely without the manufacturer’s involvement. It’s paid off. By contrast, the ARM routers that I’ve purchases lost support when they become EOL, and some of them weren’t even adequately supported before EOL.
I know this isn’t news to you of all people, but it’s been a reoccurring theme across several different ARM product lines. I wish this weren’t the case. “Fool me once, shame on you, fool me twice shame on me”
$35 is probably doable in the UK (made here), China (made there) and the US (lower sales taxes) but expectations have increased as have specs. That won’t come for free and RPi isn’t immune from inflationary forces besetting every aspect of industry. Maybe dropping the RAM, PCIe lane and USB 3.0 would get you back to that magic price point but then again, an RPi 3b might be all you need, anyway.
With regard to demand, Upton in a recent YT interview said that production of units is not the bottleneck – it’s distribution once the units leave the factory. Not sure how that works in individual markets but I am lucky enough to live a two hour drive from the RPi Store in Cambridge. I was there recently and there was no shortage of all Pis in all SKUs plus the usual third party accessories.
Having been fortunate to have got an RPi 4 from eBay for a song before Covid and all hell broke loose with NAND supply issues, I don’t feel the need for an upgrade at that price point since I already have a couple of Lenovo tiny desktops, on whose territory the RPi5 is intruding. Having said that, there are also worrying videos on YT showing how to mine crypto on RPis. It’s the last thing anyone needs.
SDug,
Don’t forget that Trump administration significantly increased US import taxes and they haven’t gotten rolled back – most imports are taxed at 15% on top of (or technically underneath) the local sales taxes. US prices have gotten significantly higher in recent years.
I don’t think $35 is viable anymore either in this era. It’s true that at least the specs have gone up, but for those who just want to tinker $60 for the basic RPI5 4GB board only without accessories is quite the price hike. For me RPI was always less about low power desktop computing and more about electronics tinkering. Honestly if I wanted a low power desktop, I’m more inclined to look at one of the many affordable ready to use mini PCs.
Yeah, there is defiantly a supply and demand issue with Pi. The solutions to high demand and low supply are either
Company Increase product price
Company increases output capacity
There is a shortage of products.
We’ve seen all three outcomes at once it seems. If I were being optimistic, I’d say that with more funds, the could increase capacity faster. But realistically, it seems the price is due to increase as well as it seems like the market would bear a price increase even with a much greater supply.
> “while I’m involved in running the thing, I don’t expect people to see any change in how we do things.”
Well, that’s a nice loophole he left there.
Problem is, by now the pi’s value lies in its ecosystem.
If anyone is willing to spend hundreds of millions on this, they will want to see profits extracted at some point. There’s a gazillion copycats out there that manage the hardware side of things just as well, so the squeezing of the ecosystem will be inevitable.
As someone who had tried many of those gazillion clones, I can easily say Raspberry Pi has a unique advantage. It works, and the software is supported.
Even nvidia with their Jetson line has less than stellar support. “Fly by the night” stuff that you get from random online stores have even less shelf life. Either they won’t be supported in the next Linux version, or they won’t be supported out of the box even.
And let’s not forget the brand name recognition, and the “compute form factor” being already used in many commercial applications:
https://pallavaggarwal.in/10-products-using-raspberry-pi-compute-module/
I don’t think they would switch to a different vendor just because “it is cheap” or “it has more RAM”. Stability is very important in these applications.
The only major threat RPi has is their (lack of) manufacturing capacity. They can never provide enough chips, and this has actually led some to look for alternatives.
sukru,
I agree with both yourself and Mote, you can find other hardware SBCs (and I own some of them, orange pi, odroids), but where RPI stands apart is community & software ecosystem. Ideally things would be like x86, where you can get x86 hardware from any vendor and boot the x86 distro that pleases you, but ARM being ARM, hardware and OS are more tightly coupled than x86.
Haha, I just posted this in another response. I agree though, the raw specs aren’t the end of the story…other vendors do have good specs, but pose other unforeseen support problems. Some are even stuck with old binary kernels released for android that don’t get updates. Universal ARM standards would alleviate this. Some have promise, but until it becomes adopted into commodity hardware we’re in a free for all and unification remains a distant dream.
It’s exactly these kinds of boards that are affected by the Linux kernel’s recent change in LTS policy.
The clones (at best) will only have a supported kernel for 2 years. And that’s assuming you get the 0-day kernel! In terms of putting anything Internet-facing, it’s basically a no-go.
Its basically why, however tempting they look on paper, I steer clear.
Adurbe,
I don’t think this is a Linux, problem, but rather an ARM platform problem. I think this summarizes it well:
It even hits Android devices, that is why Google was working on a project to make their Linux at least a little bit more modular:
https://android-developers.googleblog.com/2017/05/here-comes-treble-modular-base-for.html
The root of the problem is many electronics manufacturers do not give value to longevity. They would write a hacky driver, build it as a binary, and release it for a single specific Linux kernel. Not a kernel version, a specific kernel binary.
Even changing kernel config on the same version, and rebuilding it usually breaks them.
That is why having open source drivers, part of the Linux tree, is important, but I would estimate 99% of the time new devices don’t even bother.
Could somebody explain to me what is the point in selling and buying “non profit” companies shares?
Where’s the profit to be made here? An expectation the company will switch to for profit operations in the future?
dsmogor,
“It depends” is the correct answer. Many non-profits also wholly own for-profit subsidiaries. Like a school owning a cafeteria, or OpenAI Foundation owning the for-profit arm.
The largest example is probably IKEA. They are non-profit doing research on interior design (I think) that happens to have some factories and outlet stores. And they also license the logo from another entity (at probably a very steep fee). This time most likely all for tax purposes.
https://en.wikipedia.org/wiki/IKEA#Corporate_structure
Two words: VR porn. This is one of the main reasons that VR headsets will receive massive funding and will become more and more efficient and compact.
xDisruptor
I think you meant to post here https://www.osnews.com/story/138477/apple-vision-pro-review-magic-until-its-not/
Anyway, gaming and porn, both areas that push technology. I can see why people might want to try it out, but is it really that much more enjoyable long term? We keep revisiting VR and 3d content, every generation wants to experience it, but then it never stays around.
The 3d glasses at the movies as as light and compact as things can get, but after a few thrills the appeal kind of wares off until the next generation wants to do it all again.
This is indeed very off topic but I agree with you.
That said, I recently went to see Migration with my kids and 3D was the only timeslot that worked for us. The glasses we got were super light and comfortable. Much better than I have encountered before. I bet I could wear them around in the real world without attention. Because they fit, the register ( lack of blurriness ) was also excellent. You do just forget about the 3D effect most of the time but there were a couple of scenes where it added to the movie.
Anyway, it is the first time that I have seen a 3D movie were I thought to myself that maybe it was time to start doing more movies in 3D or even making it a normal thing. At least, it would add a little something to the theater experience.
Then again, if I could have saved a couple of dollars per person on the movie without 3D, I would have taken it.