Am I crazy, or has pine64 actually managed to build a sustainable ARM Linux company? There have been a hundred of these projects over the years. I feel like they tend to launch with a lot of fanfare, and eventually lose steam and die.
But pine64 seems to have started humbly and slowly be building momentum.
Their margins have to be tiny though, judging by the impressive build quality of my $150 PinePhone.
Yes, sounds like a miracle, but I'd say this is not something exceptional for a seasoned electronics company, which 99% of previous contenders weren't.
Building hype, and getting people's moneys is an an easy thing.
But I say it is completely impossible for newcomers to the industry to succeed commercially in anything, but a one off project carried at a very slow pace.
I've met countless people people coming from the other side of the Pacific to Shenzhen to do manufacturing, thinking that they knew what they are doing solely because they did few years of PCB design, EE, or worked a year in a small EMS in America.
All of them were up for a very rude awakening. The scope of knowledge one gets from "few years EE" in America would count for just few percents of expertise needed to run manufacturing commercially.
Just one thing that I think sinks half of them: near zero supply chain expertise. I'd say you absolutely cannot do any manufacturing in serious commercial volumes without at least 10 engineers hired just to do supply chain management full time.
If you are small manufacturer, or OEM, hunting for parts, and components, studying, and testing them, spamming Alibaba, and spending time on the phone with suppliers would be easily taking more than a half of your engineering manpower.
> ...near zero supply chain expertise. I'd say you absolutely cannot do any manufacturing in serious commercial volumes without at least 10 engineers hired just to do supply chain management full time.
As a small manufacturer with less than 10 engineers total (though in industrial automation, not B2C electronics), how do "real" electronics manufacturers do supy chain management?
We did a "small" run of a little machine that went from our usual quantity 1-3 units to quantity 100, and the strain it put on engineering was intense. Usually we just sent purchasing a BOM, this was something else. You have to be an engineer to comprehend the specs and requirements that engineering sends to you, you have to be in finance/management to understand tradeoffs of quality, savings in volume, and time, and you have to do grunt-work just to keep track of it all and keep communication and relationships active. The result was that design engineers were spending a ton of time on supply chain management long after the design had ended; do bigger shops have engineers who spend all day on the phone asking "where's my stuff" or surfing Alibaba?
> do bigger shops have engineers who spend all day on the phone asking "where's my stuff" or surfing Alibaba?
Yes, lots, and lots of them.
Can testify of working in an engineering team of 50 people, with close to half of engineering manpower spent on SCM.
> As a small manufacturer with less than 10 engineers total (though in industrial automation, not B2C electronics), how do "real" electronics manufacturers do supy chain management?
It actually becomes easier the bigger you get because you can simply buy more parts right away, in bigger batches, keep bigger inventories, and get higher part availability because suppliers themselves will be chasing after you with hopes of selling more parts.
But for everybody else, it's tough life, and 50%+ of engineering resource spend, no trick around it.
I think all of biggest ones like Foxconn, Flextronics, and Pigatron have their own proprietary IT systems for supply chain. Instead of running after suppliers, they force them themselve to enter their parts data into their system, and do test, and validation on their own too.
> I'd say you absolutely cannot do any manufacturing in serious commercial volumes without at least 10 engineers hired just to do supply chain management full time.
This is why the sole EE at your hardware startup practices Digi-Key oriented design. If a part isn't available regularly and in reasonable quantities on Digi-Key, it doesn't go into the product. Your margins will be lower than ideal, but you can usually build in batches of ~1000 this way without getting totally sunk.
Of course, if you're even moderately successful, you'll eventually find Chinese clones selling for half price, so you'd better have some other competitive advantage if you plan on taking this route.
The big difference between Pine64 and the other (phone) projects that have come and gone over the years is that Pine64 really seems to have the manufacturing (including sourcing and logistics) side down. The other projects came across as seeming to think of all the 'physical stuff' as a distraction that they'd just outsource, when in fact it's the most important part. They also don't seem to have any 'year of the Linux on X' delusions: they are in the 10's of thousands of units on PinePhone and they're fine with that.
I've been happy enough with my Pinephone (just don't call it a daily driver!) that I'm seriously considering picking up their next gen Pinebook Pro whenever that gets announced/released.
Agreed! I really like how they’re not trying to act like their products are going to revolutionize computing and destroy Apple — they’re just “we made some cool devices for hobbyists that are totally open, have fun”
Plus they’re frank that their affordable price is community orientated, and if you’re looking for a polished experience or would chargeback due to “a couple dead pixels”, please reconsider your purchase.
Looking at crunchbase, they've only raised funds using a kickstarter once. It seems like they're making enough to pay salaries and keep the show going. I know some of their products are not priced for profit, but the company as a whole seems to be doing well and growing. It doesn't seem like the typical overhyped SV company that only survives on VC money.
Considering Apple doesn't sell any devices running Linux, I don't see how they could've possibly beaten PINE64 in building a sustainable ARM Linux company.
In theory yes, in practice, only under virtualization (though I'm one of https://www.patreon.com/marcan sponsors so maybe one day this will change).
However this is a moot point as Apple doesn't sell what Pine does - Linux device. Apple sell Apple devices. If they incidentally runs Linux is irrelevant to Apple.
Unless I am doing something very wrong (and sometimes I do), most of my Python development flows fit very comfortably on a 4GB Celeron laptop (that's currently with my 8yo daughter).
If they can't perform decently on this, I need to fix it.
I live in Ireland and I really enjoy that small Celeron I mentioned in the other message - the one my daughter is using. It's small, the battery lasts for at least 5 hours, completely silent, and, while it certainly isn't as fast as my Mac or the Dell (or the Xeon under my desk), it's fast enough for my tests to run while I think about the next steps. And it did cost me about $200. It's an expendable computer.
I assume a Pinebook Pro is a much nicer experience.
Gladly? I am happy I get paid money to develop in Swift ona Apple... Happy to develop (in Python, reluctantly, nothing is perfect!) on my Raspberry PI.
We are a grumpy opinionated lot - developers. Very hard to make us glad
I totally agree with you, but would add that some of us find dealing with Apple extremely frustrating. The most valuable company in the world, and a money grubbing pack of unethical rotters...
But pine64 seems to have started humbly and slowly be building momentum.
Their margins have to be tiny though, judging by the impressive build quality of my $150 PinePhone.