Raspberry Pi CEO Eben Upton Discusses Stock Updates, Industry Prioritization
tomshardware.com
tomshardware.com
Upton claims that these companies had to use Raspberry Pi…. sounds false to me - lots of other companies changed their designs to handle the shortage.
If anything, Upton should have backed kids and schools above all - they are the one who really needed hobby computing during Covid lockdowns. Upton's justifications don’t ring true.
Also the company has treated hobbyists, kids and schools like garbage.
Buy something else.
Edit: here's a recent Ask HN: https://news.ycombinator.com/item?id=34976965
From the linked Ask HN.
Hardware is irrelevant if the software isn't comparable. I'd like to be proven wrong but if nothing has changed in the last years there are no real alternatives to the Pi.
Ex lease mini desktop Dell/Hp/Lenovo Intel machines off eBay can be good, and first class Linux support, and cheaper than raspberry pi.
Some applications can be replaced with a microcontroller like esp32 which has excellent software support.
Many of the machines above have pretty average Linux support but it seems to be improving, depends on your application.
If you mean "gives me a standard linux distro I can install software on that can talk to the network and maybe show a UI on a screen", pretty much anything works (and a lot of things people do with Pis are just that - running things like network-wide adblockers, 3D printer servers, small python scripts for whatever).
If it's somewhere in between, it again depends. E.g. if I wanted a MIPI camera, I'd probably track down a RPi and use that unless I could find someone documenting a precise setup on something else.
BeaglePlay, the newest version of the BeagleBoard:
> https://beagleboard.org/play
HN:
They're available on Amazon, 2GB board is currently $35. Same footprint as the Pi, fits most of the same cases, hats, etc. Decent support forums and community, and the images aren't too esoteric.
I have a raspi running pretty much non-stop since 2011 and am still able to update its software 12 years after purchase. All "somethings elses" which I bought had 3-4 years of software support and then their usability diminished drastically.
That these devices are useful for hobbyists, doesn't automatically mean that people want to periodically spend hours to have them working with up-to-date software.
The reason is that Linux has no stable in-kernel driver interface, so the drivers often have to change with new kernel release.
Feel free to complain on the LKML why you consider this to be a bad idea ...
Part of the problem here is that SoC drivers are hacked on top of the kernel (and are not merged in-tree). So SoC vendors have this pile of legacy and hacked-on code they need to cherry-pick every time the Kernel changes an interface they used (which is a lot) [0]. The code is also, let's be honest here, low quality and full of workarounds for implementation details (some radio-modem talking over I2C to transmit voice to the SoC???) so there's no way it would get merged into the kernel.
Of course, that costs and fortune and hampers the sales of newer chipsets so it’s not in the SoC vendor’s interest to do so.
[0] https://en.wikipedia.org/wiki/Application_binary_interface
On the other hand, it would be in a lot of vendors' cost-saving interests to massively lobby for a (much more) stable in-kernel driver interface in the Linux kernel so that the vendors only rarely have to change their Linux drivers when a new kernel version gets released.
That’s a new one for me. I would find it a lot less condescending to be told a straight “this made us the most money” reasoning. Anything else is obviously BS or incompetence.
I really hope the likes of the VisionFive SBC see decent adoption. Their reference manual seems good - but I haven’t really played around with the board all that much.
Sure, if;
- You don't run any other OS then Linux. For example Raspberry PI has pretty decent FreeBSD support. Other boards often don't support FreeBSD or not all hardware on the board has drivers.
- You don't have a custom case which relies on the design of the PI.
- You don't already have dozens or hundreds of PI's in the field and are not that enthusiastic to add a new board to the mix.
- You will probably also have to add a new board to the mix every 3 or 4 years or so since these other boards often have a lot shorter lifecycle.
- etc.
> Upton then explains that OEM customers are mainly 3,000 boards per year, with between one to ten employees who build some specific product using the Raspberry Pi and cannot use an alternative SBC. If these customers do not have access to Raspberry Pi, then they are likely to "go to the wall". Upton then explains "So the real question, when people are saying, 'Should you do this prioritization call?' is 'Should I zero some of those customers or constrain them so heavily that they go out of business or take very serious damage?'. It didn't feel like the moral thing to do."
> Upton then explains that it felt like Raspberry Pi were doing the right thing, or hoped that community members would give them the benefit of the doubt. But Upton is aware that some members will be permanently disgruntled. Upton clarifies the misconceptions that Raspberry Pi are not making boards, they are, they are just not making enough. The second misconception being that Raspberry Pi are favoring large customers (Upton uses IBM as an example) over the communtiy, when in reality it is the smaller OEM customers who are the median customer base. Upton concludes by saying that they are looking forward to now making that judgment call anymore.
Okay, yeah, they just permanently lost me. The idea that they had to make this decision to save the jobs of all those people is ridiculous.
Morally I had to support the corporation at that point.
The top 2 uses of hobby Raspberry Pis is to sit in a drawer, or play tv shows you downloaded off the internet, that's the easiest market to switch to an alternative.
If that's true, I wonder why the market hasn't corrected. With the prices Pis have had lately, I would expect them to go from drawers to ebay quickly.
My personal explanations (I do believe there is a grain of truth in each of them, but you have to take them all together to get an explanation):
- A lot of the RPi boards that lie around in the drawers are older versions in which many buyers are less interested.
- Many buyers are hesitant to buy used RPis because something might be broken with them; this risk (and attempting to get a refund) is not worth the trouble.
- In countries where there are complicated taxation laws and strong(er) customer protection laws, many hobbyists do not consider it to be worth the (felt) potential legal trouble to put their RPis on eBay.
I don't think they've talked about what their specific constraint is, but given their SoC supply afaik is still entirely dependent on Broadcom I wouldn't be surprised if Broadcom basically said "ok, you can have X units this quarter" and that's that. Not like RPi can easily replace Broadcom and not like Broadcom doesn't have demand from other potentially more important customers to meet.
A potential mitigation that the Raspberry Pi Foundation could have done: commercial users can get prioritized RPi boards, but they have to pay an insane mark-up for this priotization (say 200 € instead of 50 € for a board).
This way, corporations for which otherwise the production chains would stand still (i.e. much larger financial losses) can get a solution, but a solution that is not cheap.
This solution should appease many hobbyists, since this way they don't feel that commercial users "buy away the RPis, which were originally developed for hobbyists and schools", but rather that such commercial users rather subsidize the RPis for the hobbysist and schools via the "priotization mark-ups".
EDIT: Under https://news.ycombinator.com/item?id=35999824 cinntaile posted a related consideration.
I do believe there exist further points that make the Raspberry Pi boards attractive to companies (in opposite to SBC offers by many other companies):
c) There exist a lot of people (potential empoloyees) who have already worked with a RPi (even if for private projects) and thus know RPi's insides rather well.
d) The RPi boards (even very old ones) are supported by the manufacturer for a very long time with GNU/Linux support that (mostly) "simply works".
e) The Raspberry Pi Foundation is a British company (in opposite to boards by Chinese producers)
Potentially:
f) The esteemed Raspberry Pi brand might also give some martketing prestige to an offering of a company that uses a RPi board as a central part ("Our product uses a Raspberry Pi").
Fact check on the two Italian resellers linked from Raspberry's site
https://www.kubii.com/it/40-le-schede-raspberry-pi they have the Zero W, the 3A+ Pis are out of stock
https://www.melopero.com/categoria-prodotto/raspberry-pi/boa... same as above
That is the purpose of the Raspberry Pi Foundation, the registered charity in the UK.
I'm not sure how that fits with having so many non-educational customers.
https://register-of-charities.charitycommission.gov.uk/chari...
I do have a few boards that are faster than a Pi 4, such as the Radxa Rock5 with the Rockchip RK3588, but they cost way more than a Raspberry Pi (once again if you can find at MSRP).
Its point is that its it costs less than $20 to have a working linux machine with GPIO. Not only that, it has all the documentation to allow most people to get from 0-flashing lights.
It was and is an educational platform.
the same with the rp-2040. Its a cheap, lowish power device that easy to program, has great documentation and is versatile.
If you're trying to run docker on it in production, then you have the wrong platform.
The Pico is a completely different system and use case.
Whereas if you have a complete product designed around an SBC, respinning that is expensive, and a few hundred to a few thousand units is an ugly point where that leads to really high per-unit costs.
Whereas if you make millions of a thing, the overall development cost is similar, but spreads over many more units. (EDIT: you also probably aren't using an SBC, for similar-ish reasons: a customized board is saving you money per-unit, and millions is enough to amortize the up-front cost of that and still come out ahead)
So the economic perspective IMHO makes a lot of sense.
Of course you also have large companies making low-volume products using Pis, where I guess one could argue more that those companies should be able to take the hit of temporarily not being able to make as many as they'd like. But for smaller companies without large buffers or other business, its a big problem.
I'd be curious to what degree they've looked into the exact cases when deciding who to support.
imo the failure to ramp up production is the unforgivable part
So, you're a 5 person company that makes a widget for some specialised industry.
over the last 4 years you've spent 30% of your turnover building an enclosure, getting the moulds made, and have orders for 3k devices.
They need to be delivered inside 3 months.
A new enclosure mould will cost 5 months turnover to produce, and god knows how many man months to prototype.
The software, sensor package and PCB will need to be re-designed, certified and tested from scratch.
bankruptcy.
I'm inclined to say that that's true, and even that it might make sense in terms of money for the RPi folks to prioritize businesses for that reason, but it's worth remembering that it cuts both ways: Individuals also remember which vendor screwed them over.
However, I suspect that individuals are more likely to tolerate delays in hardware acquisition, and also more likely to stick with an RPi due to the ecosystem offering vastly greater levels of support and existing solutions.
(When I struggled to lay my hands on a Pi Zero a while back, I looked for other options, but found various negatives for the potential alternatives.)
The CEO of Raspberry Pi disagrees with you.
The truth is that browsing through alibaba proves him wrong, fast.
All he has is a trademark (which is extremely valuable, don't get me wrong).
Where can I find a SBC on Alibaba that uses the same SoC as some Raspberry Pi?
Things seem to be looking up, though - they actually have Pi 400 kits in stock, something I haven't seen in seemingly ages. If you can cannibalize a 400 or use it as-is, that's a nice option, particularly with the heat sink. Unfortunately, it's kits only, not the bare machine.
Additionally, unfortunately there does not exist a version of the Raspberry Pi 400 with 8 GiB of RAM.