This prompted me to look on Ebay, where Pi 4 prices are mad (>£100!) but the previous generations seem to be more normal. Wonder if it's worth me digging up any 3s I have and selling them.
I wanted one for a thing some months ago, and the 4s were out of stock most places and silly prices where they were available. 3s were in a similar state. I had a 3 spare but wanted the extra spec (USB3 and a bit more CPU umpf). I ended up getting a 400 instead as they were readily available at the time. When the 4s become sensibly available I'll swap it out for one and the 400 will become a toy for other purposes.
EDIT: sweet Jesus. Just checked the current state on Amazon and the 400s are listed at £117 for just the base unit, more than the £91 I paid for the full kit (the only in-stock option at the time) six months ago, the 4Gb 4 is £173 (next day prime) or £155 (up to two week delivery), and the 8Gb is down for a whopping £290! (next day, no cheaper slower options). The only Pi3 showing as in-stock ATM is £100. Maybe I should hunt out my spare 3 and make some drinking money from it!
I sold my 3B for a price that honestly makes me feel a little guilty at the height of the shortage, but to be fair it was an auction and I only set the reserve to MSRP. At least it went to someone who presumably really needed it and it was just sitting on my shelf gathering dust after being used for a project.
What worked for me was installing Twitter to get notifications from @rpilocator. Adafruit restocks (IIRC) on Wednesday or Thursday afternoons and you have a couple minute window before they sell out. Be sure to have two-factor auth set up in your Adafruit account in advance, it's required for pi orders.
That's encouraging. But we're apparently not out of the woods yet. It wasn't very long ago (3 weeks maybe?) that I was on Mouser.com (or maybe it was Newark, whatever) looking for DSP's and there were basically none to be had.
Hopefully we'll turn the corner soon and availability of most IC's will get back to normal. fingers crossed
Honestly, it depends on the kind of component. It's easier to get CPU that support android than more lambda one for example, and that's without speaking about sub component. This lead to strange situations to say the least (partial bom delivery for example, so no power, no ethernet, 2G instead of 3G/4G modem...).
Unfortunately we might not be out of the woods yet. I have an acquaintance who works in the defense industry and they told me they are having semiconductor supply chain issues because of a shortage of noble gases needed for semiconductor production. Apparently Ukraine is a huge supplier and the war has really constrained supplies.
> one large industrial conglomerate had resorted to buying washing machines just to scavenge the chips inside them for its products
Is this really more efficient than say, contacting the washing machine manufacturer and buying up the chips at a premium (i.e. above the washing machine cost).
I usually applaud hacky solutions, but this just feels like a ton of unnecessary waste. I would prefer scrapping features (like Tesla did with the USB ports) over trashing a ton of washing machines.
Yeah, in this case the manufacturer probably didn't have any extra chips. Also, in my experience, these companies were unwilling to go outside their system and charge a markup for their own stock.
Instead of giving me the ICs directly, I had to wait for them to ship an order of dev boards to Digikey that I'd already purchased and they were building to order.
> buying up the chips at a premium (i.e. above the washing machine cost).
Industries often act illogically in cases like this. You see it with many types of goods ordered a long time in advance.
If there is a sudden shortage, the people who have ordered their raw materials a long time in advance (who have effectively purchased a future for that commodity) think they made a good move and see it as a business advantage that they can continue production while their competitors who failed to plan cannot.
However, they often fail to see that it is an even better business move to still shut down production and sell those raw materials onwards to someone else who will pay handsomely for them.
Basically, the futures market should never determine if a factory is operating or not. If it does, then someone has made an irrational decision.
This is one reason I think most contracts for goods for future delivery should be public knowledge.
Ie. "Mr Smith purchased 68 tons of steel for delivery in New York June next year for $200k".
Then anyone wanting those same goods and willing to pay more can contact Mr Smith and make an offer to buy the contract.
I believe such a system would make the economy substantially more efficient, at the cost of privacy. More knowledge makes a better market. And more knowledge also increases tax revenues because under/overdeclaring goods values becomes much harder.
Why would anybody want to trade with Mr. Smith's contract in particular when a market for the contract still exists, with market makers and other participants providing bids and offers that is better what Mr. Smith can do? At least, assuming you are talking about a regulated commodities futures contracts like those which trade on the CME.
Mr Smith has a real for delivery contract, not a more common contract to own for some amount of time but no clue who is on either end. As such if you buy from Mr Smith you get real goods, while from someone else you may discover that you created a short squeeze, and not actual material is available: they may have to pay you a ton of $$$ to get out of the contract, but you don't actually get anything delivered.
If you really want the actual good and not the dollars, then you should by buying contracts backed by real goods that cannot be shorted. By contrast if you just want to make money there is more money in buying the contact that is a short with no material behind it. Of course knowing when a short squeeze will happen is left as an exercise for the reader - meaning I have no idea when this plan can work out only that once in a while it does.
If most buyers buy ahead like Mr Smith, then the liquidity in the spot market can dry up, especially in times of supply constraint.
Mr Smith and many people like him might happily sell on his steel for 10x the price he paid for it to someone desperate for some steel.
Yet, as things stand now, whenever there is severe market disruption, you frequently find goods aren't available for any price. Thats simply because Mr Smith and people like him continue to take delivery of the steel he ordered 6 months ago, and doesn't actually need right now, because he isn't aware of the guy down the street who really reeds it right away and will pay a lot.
If all goods are always 'for sale', then goods can be allocated efficiently to whoever needs them most.
Somewhere, an executive's bonus pay is based off the number of units shipped each year. Not the amount of money saved for the company. Under that kind of regime, it actually makes less sense to sell the chips at a premium price. For the executive.
> However, they often fail to see that it is an even better business move to still shut down production and sell those raw materials onwards to someone else who will pay handsomely for them.
There's also business value in being the washing machine company that didn't have to shut down production. If I'm building a lot of housing with washing machines, maybe I just wait and buy whatever just in time, but maybe I want to do a bulk buy direct from the manufacturer with delivery on a schedule; having a track record of continuing production during supply shortages makes it more likely that the deliveries will happen. If production is stopped whenever it's better to resale materials, it's hard for buyers to rely on scheduled deliveries and those buyers will turn to others.
>However, they often fail to see that it is an even better business move to still shut down production and sell those raw materials onwards to someone else who will pay handsomely for them.
Isn't this the short-term quarter-to-quarter mindset that gets constantly roasted on HN? Isn't there value to your brand in being a dependable supplier to your customers? And isn't there business value in retaining your workforce, instead of having to lay them off, since you shut the factory down?
> the futures market should never determine if a factory is operating
But what about washing machine futures?
Or more realistically, how much is shutting down the factory going to cost the washing machine manufacturer in terms of employee compensation, lost employees (more hiring and training), washing machine brand losses in the market, failure to deliver on existing contracts, and similar? They can't claim force majeure when they shut down the factory merely because they see a way to make slightly more profit than normal.
True... but that has to be compared with those same set of losses the car factory up the street will suffer if they don't get the chips...
If there is only enough chips for one factory to run, someone has to shutdown and take those losses. And the shutdown factory should be the one with less to lose, not the one who didn't buy futures in time.
> the shutdown factory should be the one with less to lose, not the one who didn't buy futures in time.
That sounds highly communist. The factory with the most to lose should have been looking out for themselves. They did not, so they suffer economically, and hopefully learn a lesson (or their bankruptcy stands as a lesson for others) for the future.
Because of their decisions, one of the factories can claim force majeure to protect them from some of the consequences of shutting down, while the other cannot, so the other shutting down and reselling their chips is probably not an option.
Salvaging chips out of washing machines might be the best possible option for everyone to fulfill their contractual obligations.
That's what you'd do in a command economy. In a market economy it's up to the company what they do; there are far more things "real people" try to maximize than just profit. You can see this as people acting irrationally, or you can see it as there're important factors your missing and/or not understanding.
Economists be like "guys, your doing it wrong - not maximising profit", economy be like "nah, your theories incomplete".
It might be more (cost) efficient to buy washing machines.
It's possible if the washing machine manufacturer knew of the real value of the chip to the buyer, they might charge extortionate pricing.
It's also possible that the total value of selling a washing machine exceeds the purchase price (e.g. how much does the washing machine manufacturer get from the detergent maker for putting a packet of detergent in the box).
Buying the machines means that you go to all their dealers and buy - thus ensuring that their dealers remain in business. When things return to normal their dealers are still able to sell to everyone else who wants a washing machine and so they don't have to build a new dealer network. In short it is probably penny wise, pound foolish to not make washing machines even though they could sell the parts for more money now.
When I tried to build a mattress, I was unable to buy certain components directly from companies, no matter which ones I contacted. Had to buy mattress toppers and rip the parts out. Guess this applies to corporations too lol
There is a lot of liability and time cost involved in a transaction. It does not necessarily make sense for businesses in the business to business space to expend effort on business to consumer sales.
I posted earlier about orders being held up because parts were unavailable. If you have a customer for your $700,000 machine that can't be shipped because a $50 module is off the market, then you bet it makes sense to buy a washing machine to extract that part. The only other alternative is to hope that there's enough margin to redesign around another, available part.
We had a client pre-COVID that found themselves in exactly this predicament. If we could have found the part they needed embedded in another machine, that's exactly the recommendation we'd have made.
The problem in these cases is not that you can't buy them at a premium - it is that you cannot buy them at all. What do you do if nobody can supply you the part at any price?
I know of certain STM32F7 SKUs being sourced for $600/piece instead of $8/piece via shady backstreet Shenzhen suppliers. The whole electronics scene really needs a transparent futures market.
Starlink provides a router with their ISP package. For the last year the router comes without an ethernet port. Wireless only. They've never said why but I assume it's because of a chip shortage. (There is an add-on for ethernet you can order for a modest price.)
ResMed started selling a version of their popular AirSense 10 AutoSet sleep apnea machine without a cellular modem. Normally the product uploads treatment data to some cloud service that ResMed then resells to your doctor. Again they haven't said explicitly but it sounds like they couldn't source the part they needed, so did without. (Hilariously the branding for this is "Card-to-Cloud", as in "you carry an SD card to your PC and upload the data manually to the cloud".)
I guess what bugs me is these companies are selling things with important features removed but aren't acknowledging they are lesser products.
>some cloud service that ResMed then resells to your doctor.
Why would they resell to the doctor, or why would the doctor be buying the info? Seems like the user should be the one charged for the data that could then be shared with their doctor. I just don't see the business sense for it being purchased by the docs unless the docs just turn around and upcharge the patient?
I should clarify I'm talking about the American system. Doctors pay for access to ResMed data. It's a convenience; otherwise the patient has to carry the SD card into the doctor's office and hope they bother to upload and analyze it. (Yes, this is stupid.)
While I'm at it, patients really interested in their sleep apnea data can use OSCAR or https://sleephq.com/ to analyze the data from the ResMed SD card themselves. These don't use the cloud service. ResMed also has a limited site for the patient themselves to see a limited version of their cloud data.
>I just don't see the business sense for it being purchased by the docs unless the docs just turn around and upcharge the patient?
Seems like you understand perfectly. The doctor is the one who can interpret the data, so they can ask their patient to come in for regular checkups and make sure everything is going great
I've commented on this before on HN. I have sleep apnea and a ResMed CPAP. A very US-centric response:
Insurance also seems to have access to this data in one way or another (not sure if direct with ResMed or via your doctor). If you don't hit your "compliance target" for treatment, insurance will cease paying 'their share' of the cost.
So I found myself in a downward spiral of
0. Struggling to sleep with sleep apnea. Tons of ramifications of not getting good sleep that I don't need to get into.
1. Struggling to adapt to sleeping with this device strapped to my face. I would take it off unconsciously.. I would often find the hose laying next to me in the morning still "pumping", or be told by my wife of belligerent (unconscious) responses to her prompting me to put it back on in the middle of the night.
2. Losing sleep from constantly waking up throughout the night because of the CPAP/from anxiety
3. Anxiety over not meeting my compliance target of 4 hours of use per night (as interpreted by the CPAP knowing the proper amount of back-pressure when in use i.e. not just turned on and not worn).
4. Getting charged more because I did not meet my targets for enough time
5. Eventually owning the CPAP outright
6. Eventually discontinuing therapy and somehow having slightly improved levels of sleep that I've never bothered to attempt to restart therapy
It took me weeks to get used to mine, but I find I do sleep better with it than without my CPAP machine. With respect to insurance, I replaced my original machine without going through insurance. It would have cost me more out of pocket for the doctor visit and on-site sleep study.
The state of health “insurance” in this country is mind-boggling. But we continue to vote for politicians that vote against our own best interests. We need more than these two f’d up parties.
I've said it in other threads, but these "two f'd up parties" are the expected end result of our electoral system[1]. While we very occasionally see one of the parties replaced with an upstart, the parties eventually return to a similar status quo. Grassroots movements to implement run-off elections, approval voting, or other similar plans are seeing some traction. If you really want to see change, searching for these groups locally (and volunteering or donating) might be worth the effort.
this is the natural evolution of a conversation. i feel no apology should be necessary. taking a conversation to new related directions is often where that Aha! or Eureka! moment is found.
personally, i find people that get upset and feel an apology is warranted are people i find quite boring to have a conversation with
Before I got my CPAP set up the one thing I found that really helped me was sleeping on an inclined pillow, up about 20 degrees. Not comfortable but helped me breathe. There are other treatments for sleep apnea if CPAP isn't working for you; some surgical options, oral appliances, fancier breathing machines. Worth talking to a sleep specialist doctor about.
Well I broke up with my long term girlfriend and complaints about my sleep apnea stopped. Not sure if this approach is viable for you, but it will solve most peoples’ sleep apnea problem (snoring annoying other people).
Other good ideas are to try one of the mouth guards or a new pillow that keeps your head/throat area sort of tight.
Sorry I can't be more helpful, as I have no good answer. I do have good periods and bad periods.
Though not massively overweight (26.5 BMI), better eating and fitness habits seemed to have helped. Still working on getting to a healthier shape. Depending on the type of apnea, this can help a good deal.
Less electronics/screens in the evening. No more streaming tv/video games until right before bed. Jury's still out on the effects of blue light etc but I really try to limit my electronics to a Kindle (Oasis with the warm color temp feature) before bed.
I admit I didn't try too hard with different masks/setups, that could help. Though I was mostly using one of the most comfortable "nasal pillow" interfaces already.
I will second another commenter's suggestion of a more elevated pillow setup. I use a regular pillow with a buckwheat hull pillow on top. Bad for my neck but I've slept that way for awhile and have adjusted, it's hard to go back to a "more optimal pillow height".
Here in Germany, "Mandibular advancement splint"[1] is a popular device recommended by doctors for less severe apnea cases. I was diagnosed with a light apnea and now I am getting this device as my test results showed that a CPAP treatment would be overkill but this could help. I am afraid I can't share more experiences as I am waiting the device to be prepared after 2 hours of dentist visit where they took some measurements.
For me, it's a combination of always ensuring my sinuses stay open and proper head placement for the mask to stay tight. Google "ayr gel" - I use the gel specifically. In dry periods or for congestion, that seems to (often) be enough to keep my sinuses open. For my current cpap mask I have to lay flat with no pillows for a proper fit. Then there's all the sleep hygiene stuff - no lights, proper bedroom temperature (cold), etc. (No lights means no lights - not even reflections from hall lights or the indicator lights from electronics can be tolerated.)
I'm so sorry you've gone through this, that sounds miserable. When you feel up for it I hope you can seek treatment again.
For context for other readers: the cost of one of these machines is about $800. That's a very small amount of money, often less than a single month's insurance premium. And for that the insurance companies turn medical treatment devices into anxiety-inducing spy machines. It's hideous.
I pay ~$500/month for "pretty good" health insurance, and my employer pays roughly the same (a "benefit" that the employer pays 'half'! /s )
And then you have to pay small fees when you go visit a doctor ($25-ish each time), pay for medications (expensive). Emergency room swipes your credit card before a doctor even sees you ($150 co-pay for my plan). Pay thousands for an MRI (that the insurance company has graciously negotiated down 95%!)
For a 30 year old it's $400 to $1100 a month depending on coverage level. A mediocre "Silver" plan is $400 at age 20 to $1060 at age 65, at which point you qualify for nationalized healthcare.
Healthcare cost, no, but monthly insurance cost, yes.
$800 per months plus out of pocket sums for each appointment is batshit insane.
Plus, loosing it when you loose your job is even more difficult. I think I finally understand why every Michael Bay movie is starting with a surprise medical bill. I always thought it was somewhat silly.
To give you a example, I pay 75€ per months for the entire family, and we don't have to pay anything when going to the doctor.
I have been having enough difficulty I've actually tried to restart in recent times. Vicious cycle of allergies and sleep deprivation make it hard to continue certain times of the year.
It was a lesson to be learned, I was unaware when starting treatment that you could buy one outright. Being a 60601 medical device I thought I had to just go along for the ride...
there is no shortage of 1000BaseT PHY and ethernet chips or you would see absolute chaos in the market of $50 to $150 home/small office routers and ethernet switches. The sort of all-in-one CPU/RAM/ethernet chip that's built into a small router like the starlink one already implements ethernet it just needs the PHY to do it
If mikrotik and ubiquiti can still source the chips to build $50 routers, which is the case, there is no extreme shortage that spacex could not handle through its own ability to find manufacturers and vendors.
The starlink v2 terminal coming with a weird proprietary copper cable interface and wifi-only-to-clients is for entirely other weird spacex related reasons.
edit: here is a photo of the starlink proprietary-connector-to-1000BaseT adapter, there is not a whole lot going on here.
Well there is shortage of PHY chips. Decent TI chips go for 50$ a piece over broker. Minimum order quantity 200 pieces. We took ancient Realtek parts for 7$ from broker. Normally it’s <4$. These are sad news from last 2 months. Regular distribution channels have zero stock with funny lead times of 1,5 years.
At the quantities and price constraints that spacex is manufacturing the low cost home router/poe injector for the v2 and v3 standard model starlink terminals, however, i don't think they have any difference in purchasing power or use of low cost PHY compared to what you'd find in like a $50 mikrotik:
This article is out of date. We have chip glut situation right now, both Intel and NVIDIA have pushed back their 4nm orders. And 7/8nm already produced chips are piling up with no buyer to be found (especially GPUs).
So many smartphones are thrown away in the trash each year. How many chips can be scavenged from them?
Perhaps chip shortage will finally put an end to the planned obsolesce of the smartphone industry, and bring some kind of stability to the whole thing.
Early in my EE career at a very small electronics manufacturer, "pulls" (cheap components that were removed from equipment) were a thing. Lots of vendors in the back pages of the Component Shopper (I think that was the name) rag advertised them.
As I moved on to companies where supply chain quality was important, we didn't do that anymore, but I'd be surprised if that marketplace has gone away. Especially for hard to find parts. I know of companies where 6-figure orders were being held up because of the inability to find a few $5 components.
It's less likely to happen for consumer products that are price-sensitive, though. There's a big difference between doing that with low-integration DIP parts vs fine-pitch SMT devices that have orders of magnitude higher transistor count.
This is a whole industry in China. There are buildings in Shenzhen with hundreds of vendors selling scavenged storage and memory chips, because those are highly standardized. And discontinued chips from the 80s and 90s are salvaged and sold on ebay and AliExpress so people can repair their old electronics. I've even seen broken laptops get their CPUs desoldered and turned into desktops with custom adapter boards. They're incredibly creative when there's profit to be made.
Fun fact: this industry also has a habit of fraudulently remarking harvested chips (which are usually fine anyway) to look like brand-new ones. It caused a bit of a minor scandal in the retrocomputing community when this was found out.
> So many smartphones are thrown away in the trash each year. How many chips can be scavenged from them?
Zero!? As far as I know, chips are inherently disposable products, recyling chips is almost an impossibility in modern electronics due to reliability problems. Nearly all chip datasheets contain the warning: DO NOT solder more than twice, which is the number of passes it takes to reflow-solder the front and back side of the PCB. And even this is only possible under carefully controlled reflow oven temperature profile, the max time interval during peak temperature should be less than 10 seconds. Any extra soldering steps or heat voids all warranty, since it exposes the chip under additional thermal stresses, and may damage the chip or degrade its long-term reliability. To recycle a chip, it takes two additional passes, first to remove it from the motherboard, next to re-tin or re-ball the chips, finally to install it on new boards, by now the chip would've experienced 6 thermal-stress events.
Consider the fact that modern surface-mount technology has progressed to a point when even reliably assembling circuit boards using brand-new chips can be technically challenging for high yield - in mass production, it takes lots of tweakings, such as solder stencil opening, solder alloy selections, solderability test of the tinning on components, reflow oven temperature profile, inert gas protection, X-ray inspection, and even Scanning Electron Microscopes to perform metallurgical analysis for troubleshooting - I fail to see how recycling is ever going to work. It adds a huge number of additional uncertainties.
Sure, repair shops and R&D labs do it all the time, and it works at a small scale. But reliably doing it at an industrial scale will be extremely problematic. This is why even a functional "counterfeit" chip (salvaged and remarked) is a great cause of concern, these components have been heavily abused, and nobody should ever trust it in a piece of critical equipment. Are you going to trust your life on a car when the microcontroller chip in the ECU is a salvaged part? Or pay $700 for a smartphone if a chip has a time-to-failure of who-knows-how-long hours?
Even back in the 80s when all the chips were through-hole components, the reliability of reused components was already a problem. In the days of QFN, BGA, LGA, CSP and WLP chips, I don't see any solution at all. This is the dark side of the semiconductor industry - nearly all chips are disposable.
Nevertheless, this is a one-sided perspective from circuit board assemblers, which is what I'm familiar with. I don't know anything about electronics recycling. Thus, if there's any progress and innovation in the field of recycling, I'd love to know. Perhaps the manufacturers are overcautious on reliability and it's not really that bad?
It's time to stop treating hardware like a toy or "the newest thing" and start thinking about ways to extend and expand use. Not just reuse as intended but different uses. As hardware gets more powerful and advanced, it's time software and the OS stared getting more versatile as well.
What we really need is advanced work on local asymmetrical interconnected systems.
I follow news from the boutique modular synthesizer space, and it has been a BLOODBATH lately.
The legendary OP-1 was recently discontinued. Prices are going through the roof. Multiple popular modular synth companies have closed up shop recently due to the chip shortage.
The OP-1 is a bad example. It was discontinued because they replaced it with the OP-1 field - essentially the same thing but with a few feature tweaks.
But the bloodbath is real. Companies are shutting down (WMD, synthesizers.com), delaying or killing products (bastl, dirtywave, behringer), and raising prices (moog).
For the sake of the little guys who can't weather the storm much longer, I really hope this lets up asap!
companies using chips from washing machines looks fishy / fake to me, wont they have some kind of security so those chips cant be re flashed, very naive to think that we can de solder and re flash the chip as if its brand new.
83 comments
[ 2.8 ms ] story [ 83.1 ms ] threadEDIT: sweet Jesus. Just checked the current state on Amazon and the 400s are listed at £117 for just the base unit, more than the £91 I paid for the full kit (the only in-stock option at the time) six months ago, the 4Gb 4 is £173 (next day prime) or £155 (up to two week delivery), and the 8Gb is down for a whopping £290! (next day, no cheaper slower options). The only Pi3 showing as in-stock ATM is £100. Maybe I should hunt out my spare 3 and make some drinking money from it!
I've not checked other sources.
The other a bit less common one is only few months tho, compared to year+ so I guess it slowly is. Inflation probably helped too
Small end-users receiving chips is a lagging indicator of the state of the supply chain. Companies are already warning about surplus inventory, ie:
https://moneyweek.com/economy/global-economy/605232/semicond...
Hopefully we'll turn the corner soon and availability of most IC's will get back to normal. fingers crossed
Is this really more efficient than say, contacting the washing machine manufacturer and buying up the chips at a premium (i.e. above the washing machine cost).
I usually applaud hacky solutions, but this just feels like a ton of unnecessary waste. I would prefer scrapping features (like Tesla did with the USB ports) over trashing a ton of washing machines.
Instead of giving me the ICs directly, I had to wait for them to ship an order of dev boards to Digikey that I'd already purchased and they were building to order.
Industries often act illogically in cases like this. You see it with many types of goods ordered a long time in advance.
If there is a sudden shortage, the people who have ordered their raw materials a long time in advance (who have effectively purchased a future for that commodity) think they made a good move and see it as a business advantage that they can continue production while their competitors who failed to plan cannot.
However, they often fail to see that it is an even better business move to still shut down production and sell those raw materials onwards to someone else who will pay handsomely for them.
Basically, the futures market should never determine if a factory is operating or not. If it does, then someone has made an irrational decision.
Ie. "Mr Smith purchased 68 tons of steel for delivery in New York June next year for $200k".
Then anyone wanting those same goods and willing to pay more can contact Mr Smith and make an offer to buy the contract.
I believe such a system would make the economy substantially more efficient, at the cost of privacy. More knowledge makes a better market. And more knowledge also increases tax revenues because under/overdeclaring goods values becomes much harder.
If you really want the actual good and not the dollars, then you should by buying contracts backed by real goods that cannot be shorted. By contrast if you just want to make money there is more money in buying the contact that is a short with no material behind it. Of course knowing when a short squeeze will happen is left as an exercise for the reader - meaning I have no idea when this plan can work out only that once in a while it does.
Mr Smith and many people like him might happily sell on his steel for 10x the price he paid for it to someone desperate for some steel.
Yet, as things stand now, whenever there is severe market disruption, you frequently find goods aren't available for any price. Thats simply because Mr Smith and people like him continue to take delivery of the steel he ordered 6 months ago, and doesn't actually need right now, because he isn't aware of the guy down the street who really reeds it right away and will pay a lot.
If all goods are always 'for sale', then goods can be allocated efficiently to whoever needs them most.
There's also business value in being the washing machine company that didn't have to shut down production. If I'm building a lot of housing with washing machines, maybe I just wait and buy whatever just in time, but maybe I want to do a bulk buy direct from the manufacturer with delivery on a schedule; having a track record of continuing production during supply shortages makes it more likely that the deliveries will happen. If production is stopped whenever it's better to resale materials, it's hard for buyers to rely on scheduled deliveries and those buyers will turn to others.
Isn't this the short-term quarter-to-quarter mindset that gets constantly roasted on HN? Isn't there value to your brand in being a dependable supplier to your customers? And isn't there business value in retaining your workforce, instead of having to lay them off, since you shut the factory down?
But what about washing machine futures?
Or more realistically, how much is shutting down the factory going to cost the washing machine manufacturer in terms of employee compensation, lost employees (more hiring and training), washing machine brand losses in the market, failure to deliver on existing contracts, and similar? They can't claim force majeure when they shut down the factory merely because they see a way to make slightly more profit than normal.
If there is only enough chips for one factory to run, someone has to shutdown and take those losses. And the shutdown factory should be the one with less to lose, not the one who didn't buy futures in time.
That sounds highly communist. The factory with the most to lose should have been looking out for themselves. They did not, so they suffer economically, and hopefully learn a lesson (or their bankruptcy stands as a lesson for others) for the future.
Because of their decisions, one of the factories can claim force majeure to protect them from some of the consequences of shutting down, while the other cannot, so the other shutting down and reselling their chips is probably not an option.
Salvaging chips out of washing machines might be the best possible option for everyone to fulfill their contractual obligations.
Economists be like "guys, your doing it wrong - not maximising profit", economy be like "nah, your theories incomplete".
It's possible if the washing machine manufacturer knew of the real value of the chip to the buyer, they might charge extortionate pricing.
It's also possible that the total value of selling a washing machine exceeds the purchase price (e.g. how much does the washing machine manufacturer get from the detergent maker for putting a packet of detergent in the box).
I posted earlier about orders being held up because parts were unavailable. If you have a customer for your $700,000 machine that can't be shipped because a $50 module is off the market, then you bet it makes sense to buy a washing machine to extract that part. The only other alternative is to hope that there's enough margin to redesign around another, available part.
We had a client pre-COVID that found themselves in exactly this predicament. If we could have found the part they needed embedded in another machine, that's exactly the recommendation we'd have made.
[edited]
ResMed started selling a version of their popular AirSense 10 AutoSet sleep apnea machine without a cellular modem. Normally the product uploads treatment data to some cloud service that ResMed then resells to your doctor. Again they haven't said explicitly but it sounds like they couldn't source the part they needed, so did without. (Hilariously the branding for this is "Card-to-Cloud", as in "you carry an SD card to your PC and upload the data manually to the cloud".)
I guess what bugs me is these companies are selling things with important features removed but aren't acknowledging they are lesser products.
Why would they resell to the doctor, or why would the doctor be buying the info? Seems like the user should be the one charged for the data that could then be shared with their doctor. I just don't see the business sense for it being purchased by the docs unless the docs just turn around and upcharge the patient?
While I'm at it, patients really interested in their sleep apnea data can use OSCAR or https://sleephq.com/ to analyze the data from the ResMed SD card themselves. These don't use the cloud service. ResMed also has a limited site for the patient themselves to see a limited version of their cloud data.
Seems like you understand perfectly. The doctor is the one who can interpret the data, so they can ask their patient to come in for regular checkups and make sure everything is going great
Insurance also seems to have access to this data in one way or another (not sure if direct with ResMed or via your doctor). If you don't hit your "compliance target" for treatment, insurance will cease paying 'their share' of the cost.
So I found myself in a downward spiral of 0. Struggling to sleep with sleep apnea. Tons of ramifications of not getting good sleep that I don't need to get into. 1. Struggling to adapt to sleeping with this device strapped to my face. I would take it off unconsciously.. I would often find the hose laying next to me in the morning still "pumping", or be told by my wife of belligerent (unconscious) responses to her prompting me to put it back on in the middle of the night. 2. Losing sleep from constantly waking up throughout the night because of the CPAP/from anxiety 3. Anxiety over not meeting my compliance target of 4 hours of use per night (as interpreted by the CPAP knowing the proper amount of back-pressure when in use i.e. not just turned on and not worn). 4. Getting charged more because I did not meet my targets for enough time 5. Eventually owning the CPAP outright 6. Eventually discontinuing therapy and somehow having slightly improved levels of sleep that I've never bothered to attempt to restart therapy
The state of health “insurance” in this country is mind-boggling. But we continue to vote for politicians that vote against our own best interests. We need more than these two f’d up parties.
I've said it in other threads, but these "two f'd up parties" are the expected end result of our electoral system[1]. While we very occasionally see one of the parties replaced with an upstart, the parties eventually return to a similar status quo. Grassroots movements to implement run-off elections, approval voting, or other similar plans are seeing some traction. If you really want to see change, searching for these groups locally (and volunteering or donating) might be worth the effort.
1 - https://en.wikipedia.org/wiki/Duverger%27s_law
(apologies about derailing into subthread)
this is the natural evolution of a conversation. i feel no apology should be necessary. taking a conversation to new related directions is often where that Aha! or Eureka! moment is found.
personally, i find people that get upset and feel an apology is warranted are people i find quite boring to have a conversation with
Before I got my CPAP set up the one thing I found that really helped me was sleeping on an inclined pillow, up about 20 degrees. Not comfortable but helped me breathe. There are other treatments for sleep apnea if CPAP isn't working for you; some surgical options, oral appliances, fancier breathing machines. Worth talking to a sleep specialist doctor about.
Other good ideas are to try one of the mouth guards or a new pillow that keeps your head/throat area sort of tight.
If we just stop testing people, the number of cases will go down.
If you turn up the radio, that weird noise the car is making will go away.
Though not massively overweight (26.5 BMI), better eating and fitness habits seemed to have helped. Still working on getting to a healthier shape. Depending on the type of apnea, this can help a good deal.
Less electronics/screens in the evening. No more streaming tv/video games until right before bed. Jury's still out on the effects of blue light etc but I really try to limit my electronics to a Kindle (Oasis with the warm color temp feature) before bed.
I admit I didn't try too hard with different masks/setups, that could help. Though I was mostly using one of the most comfortable "nasal pillow" interfaces already.
I will second another commenter's suggestion of a more elevated pillow setup. I use a regular pillow with a buckwheat hull pillow on top. Bad for my neck but I've slept that way for awhile and have adjusted, it's hard to go back to a "more optimal pillow height".
[1] https://en.wikipedia.org/wiki/Mandibular_advancement_splint
For context for other readers: the cost of one of these machines is about $800. That's a very small amount of money, often less than a single month's insurance premium. And for that the insurance companies turn medical treatment devices into anxiety-inducing spy machines. It's hideous.
Jesus fucking christ, what the fuck ?! How much does health insurance cost in the US ?
And then you have to pay small fees when you go visit a doctor ($25-ish each time), pay for medications (expensive). Emergency room swipes your credit card before a doctor even sees you ($150 co-pay for my plan). Pay thousands for an MRI (that the insurance company has graciously negotiated down 95%!)
MDSave.
I got a brain MRI for ~$450.
You can shop California's market yourself; try 94114 for San Francisco. https://www.coveredca.com/
For a 30 year old it's $400 to $1100 a month depending on coverage level. A mediocre "Silver" plan is $400 at age 20 to $1060 at age 65, at which point you qualify for nationalized healthcare.
$800 per months plus out of pocket sums for each appointment is batshit insane.
Plus, loosing it when you loose your job is even more difficult. I think I finally understand why every Michael Bay movie is starting with a surprise medical bill. I always thought it was somewhat silly.
To give you a example, I pay 75€ per months for the entire family, and we don't have to pay anything when going to the doctor.
It was a lesson to be learned, I was unaware when starting treatment that you could buy one outright. Being a 60601 medical device I thought I had to just go along for the ride...
If mikrotik and ubiquiti can still source the chips to build $50 routers, which is the case, there is no extreme shortage that spacex could not handle through its own ability to find manufacturers and vendors.
The starlink v2 terminal coming with a weird proprietary copper cable interface and wifi-only-to-clients is for entirely other weird spacex related reasons.
edit: here is a photo of the starlink proprietary-connector-to-1000BaseT adapter, there is not a whole lot going on here.
https://www.reddit.com/r/Starlink/comments/rxoz16/gen_3_ethe...
https://mikrotik.com/product/hex_s
https://viva-telecom.org/images/MIKROTIK/hex-s_6.jpg
Perhaps chip shortage will finally put an end to the planned obsolesce of the smartphone industry, and bring some kind of stability to the whole thing.
The hard part I see is the issue with finding usable drivers for them
As I moved on to companies where supply chain quality was important, we didn't do that anymore, but I'd be surprised if that marketplace has gone away. Especially for hard to find parts. I know of companies where 6-figure orders were being held up because of the inability to find a few $5 components.
It's less likely to happen for consumer products that are price-sensitive, though. There's a big difference between doing that with low-integration DIP parts vs fine-pitch SMT devices that have orders of magnitude higher transistor count.
Zero!? As far as I know, chips are inherently disposable products, recyling chips is almost an impossibility in modern electronics due to reliability problems. Nearly all chip datasheets contain the warning: DO NOT solder more than twice, which is the number of passes it takes to reflow-solder the front and back side of the PCB. And even this is only possible under carefully controlled reflow oven temperature profile, the max time interval during peak temperature should be less than 10 seconds. Any extra soldering steps or heat voids all warranty, since it exposes the chip under additional thermal stresses, and may damage the chip or degrade its long-term reliability. To recycle a chip, it takes two additional passes, first to remove it from the motherboard, next to re-tin or re-ball the chips, finally to install it on new boards, by now the chip would've experienced 6 thermal-stress events.
Consider the fact that modern surface-mount technology has progressed to a point when even reliably assembling circuit boards using brand-new chips can be technically challenging for high yield - in mass production, it takes lots of tweakings, such as solder stencil opening, solder alloy selections, solderability test of the tinning on components, reflow oven temperature profile, inert gas protection, X-ray inspection, and even Scanning Electron Microscopes to perform metallurgical analysis for troubleshooting - I fail to see how recycling is ever going to work. It adds a huge number of additional uncertainties.
Sure, repair shops and R&D labs do it all the time, and it works at a small scale. But reliably doing it at an industrial scale will be extremely problematic. This is why even a functional "counterfeit" chip (salvaged and remarked) is a great cause of concern, these components have been heavily abused, and nobody should ever trust it in a piece of critical equipment. Are you going to trust your life on a car when the microcontroller chip in the ECU is a salvaged part? Or pay $700 for a smartphone if a chip has a time-to-failure of who-knows-how-long hours?
Even back in the 80s when all the chips were through-hole components, the reliability of reused components was already a problem. In the days of QFN, BGA, LGA, CSP and WLP chips, I don't see any solution at all. This is the dark side of the semiconductor industry - nearly all chips are disposable.
Nevertheless, this is a one-sided perspective from circuit board assemblers, which is what I'm familiar with. I don't know anything about electronics recycling. Thus, if there's any progress and innovation in the field of recycling, I'd love to know. Perhaps the manufacturers are overcautious on reliability and it's not really that bad?
What we really need is advanced work on local asymmetrical interconnected systems.
The legendary OP-1 was recently discontinued. Prices are going through the roof. Multiple popular modular synth companies have closed up shop recently due to the chip shortage.
What can I say, it sucks!
But the bloodbath is real. Companies are shutting down (WMD, synthesizers.com), delaying or killing products (bastl, dirtywave, behringer), and raising prices (moog).
For the sake of the little guys who can't weather the storm much longer, I really hope this lets up asap!
https://www.synthtopia.com/content/2022/07/12/supply-chain-i...