I just checked my Amazon order history right now. Last time I bought a Samsung Evo Plus 1 TB SSD was last August, 2020, for $189. Price right now if I buy it today is $159. Last HDD I bought was a 12 TB Seagate Ironwolf for $269, in October 2020. Today's price for the same drive is $325. It is out of stock, but says I can get it May 18th.

That's a very limited dataset, no?

In 2012 there was a shortage due to a flood.

They rented a truck and hit up all Costcos (I think it was) from Florida to I don't know where buying all of the drives they could. There was a limit on how many so it was a mission to do it.

Edit to add a link and change "Around 2010" to 2012,

https://www.backblaze.com/blog/backblaze_drive_farming/

I'm not sure how pricing works where you live, but to get optimal pricing I usually have to wait a few months, or maybe even a year (black Friday). Given that all time lows for hard drive prices haven't budged much in the past few years, why buy every few months rather than stocking up when prices are low?

Anyways for my data point: there was some external hard drives on sale for 5-10% higher than "optimal" prices.

>and stock disappeared for 4 TB+.

That's strange because those drives have terrible TB/$ compared to 10+TB drives.

Opportunity cost. By keeping my cash in bitcoin I can now buy (well, not with this shortage) 5 times more hard drives than last black Friday.

> By keeping my cash in bitcoin I can now buy (well, not with this shortage) 5 times more hard drives than last black Friday.

Probably only 4 after you pay the appropriate capital gains taxes, I'd hope. :)

If they have to. Not sure how it works in Romania, but in Germany you can sell coins held for 1 year or more without having to pay taxes.

> Prices on Amazon UK are also out of control

...they are not. Sale last week for 8TB at £14/TB. I would say that prices are marginally higher than they were a few weeks ago...average prices are probably £16/TB at the top consumer end but supply is fine, this price is not particularly high historically, it is not unmanageable for individuals, SSDs prices are cratering (hint: buy stuff that doesn't have massive TBW, these consumer drives are unusuable for Chia...prices of this stuff has dropped by 20-30% in the past few months), and there is ample supply on other sites too (I believe the issue with Amazon is that they just ran some huge discounts on the external drives popular with consumers..they will be restocked in a week or so).

Also, you can pick up used stuff too. The supply of used drives with lots of life left is basically endless. And no-one is mining with 4TB drives.

My sample size of 1 shows little to no move, although I wasn't looking at drives as large as you.

In my case I was having my eyes on 5 TB 2.5" drives. I don't need them, I'd just maybe like them, so I haven't pulled the trigger, but I have them set aside on Amazon France. The price has stayed within a few euros of 110 € for the last 2-3 months.

My advice would be to decide today if you want them or not. They are not going to get cheaper, and could get significantly more expensive in days.

Same here for the same product, but in Germany.

Shortages are good.

It's the perfect time to set up manufacturing capability, so nationalists can be happy.

It's a good time to push for hire wages, workers should do it.

If there is some alternative industrial process, now's the time to try it.

This is no 70s oil moment, because there's no indication fewer widgets are being produced than prepandemic.

This doesn't change that the shortages are caused by processes with infinite appetites. The demand grows with the supply, for hardware used by cryptocurrencies. To me it looks like the polar opposite of sustainability, which is distressing.

I just don’t get the sustainability argument..

Who decided that all computing must be for a sustainable or meaningful purpose? Cryptocurrency, as wasteful as it might appear, is like a drop in the ocean compared to the data centers of Google, Facebook, Microsoft, etc. Likes on Facebook provide much less value to society than a distributed currency that circumvents central control over the financial system.

Second, Crypto doesn’t exist in a vacuum. How much “energy” does it take to disassemble a mountain and suck out precious metals or diamonds? How much energy does it take to manufacture all the coins and paper money the world over? And how many innocent people have been gunned down for the petrodollar or some other dispute over fiat currency?

> Cryptocurrency, as wasteful as it might appear, is like a drop in the ocean compared to the data centers of Google, Facebook, Microsoft, etc.

It isn't, as 'mytherin points out.

> Likes on Facebook provide much less value to society

That's absurdly untrue. For all the warts of big social media players, they still provide tremendous social value - rich long-distance communication, organizing ad-hoc groups of people (including, in particular, meatspace ones), improved exchange of news worldwide, making things easier for local commerce, ... I could go on. And while these services can, and are, being misused, they are still at least doing something useful.

> than a distributed currency that circumvents central control over the financial system.

Except it doesn't. But even if it did, this isn't necessarily a good idea in the first place.

> How much “energy” does it take to disassemble a mountain and suck out precious metals or diamonds?

Less than it takes to run Bitcoin, and it's also a one-time expense.

> How much energy does it take to manufacture all the coins and paper money the world over?

Not that much on an ongoing basis. It's also O(n) with respect to the amount of extra money you need in circulation. A trivial cost compared to what the upkeep would be if we tried to run the world economy on Bitcoin.

> And how many innocent people have been gunned down for the petrodollar or some other dispute over fiat currency?

People kill other people for profit. That's a fact of life, independent of the way money is represented. That they don't kill each other over crypto says only one thing: that right now, cryptocurrencies are completely irrelevant to the economy. If crypto ever sees any serious adoption, a proportional body count will follow.

(Also, I imagine the number for Bitcoin is already greater than zero, given that cryptocurrencies are disproportionally popular with criminals compared to general population.)

> People kill other people for profit. That's a fact of life, independent of the way money is represented. That they don't kill each other over crypto says only one thing: that right now, cryptocurrencies are completely irrelevant to the economy. If crypto ever sees any serious adoption, a proportional body count will follow.

This is a fascinating insight I had not considered.

> Not that much on an ongoing basis. It's also O(n) with respect to the amount of extra money you need in circulation.

I think this isn't quite correct, but is even more in favor of fiat currency. Minting money costs in proportion to the number of bills/coins. If more value is needed, the denomination being printed can be changed to represent more value with the same cost.

On the other hand, cryptocurrencies require wasting resources proportional to the total value represented by them. Any less and they are vulnerable to attacks. Therefore, (as I'm sure will come as a shock to nobody) this is another way that cryptocurrencies scale worse than fiat.

Supplies of most crypto currencies are finite, and the currencies appreciate in general, see bitcoin and etherium.

Supplies of most fiat currencies are effectively infinite, and the currencies depreciate in general, see virtually anything from USD to the currency of Zimbabwe.

While true, that's a bit of a non sequitur. The value represented by a currency isn't measured in units of that currency, but in terms of what can be bought with that currency. If that purchasing power goes up, then so does the incentive for launching an attack. Therefore, to maintain bitcoin's security, there must be a corresponding increase in mining costs to prevent such an attack from occurring.

Bitcoin requires energy expenditure proportional to the value represented by bitcoin. Anything less, and it is vulnerable to attack. Bitcoin fundamentally does not scale.

Very well written

Thank you.

I do think there is a lot of value in cryptocurrency research. The theoretical insights gained into ways of provably ensuring consensus in an untrusted environment, or tamper-proofing information, are already useful.

But that could all be done in a research setting (as it mostly already is) - we don't need every piece of such research to be immediately deployed in the form of a get rich quick scheme. So I'd strongly support an official, government-backed moratorium on the use of all cryptocurrencies, until researchers figure out a way to make one that isn't an exponentially growing, insatiable power sink.

We can test out the research incrementally in form of systems that don't pay people to grow chains. Like some of the project ideas I've seen thrown around big, non-financial companies - e.g. making it easier to ensure supply chain integrity. While in general I'm strongly suspicious of anything that even mentions "blockchain", some of these ideas don't look that stupid (but then, I'm not a domain expert in these areas, so the blockchain-backed solutions may all just be hot air).

>Cryptocurrency, as wasteful as it might appear, is like a drop in the ocean compared to the data centers of Google, Facebook, Microsoft, etc.

https://www.eenewseurope.com/news/first-seven-customers-5nm-...

Are you sure? Bitmain is competing for 5nm wafers against the largest electronics manufacturers in the world.

Bitcoin alone is estimated to use around 129 TWh of electricity [1], compared to 12 TWh for Google or 200 TWh for literally all other data centers [2]. Hardly a drop in the bucket, unless that drop is the size of 10 buckets. Surely not even the most hardcore cryptocurrency enthusiast would argue that crypto provides anywhere near the utility of the rest of the entire internet.

[1] https://www.visualcapitalist.com/visualizing-the-power-consu...

[2] https://www.deccanherald.com/business/why-bitcoin-uses-10-ti...

>It's the perfect time to set up manufacturing capability, so nationalists can be happy.

>If there is some alternative industrial process, now's the time to try it.

wouldn't this take years?

>It's a good time to push for hire wages, workers should do it.

This is just speculation on my part, but I suspect hard drive manufacturing is much more labor intensive than capital intensive. Therefore I doubt this will translate into higher wages.

Companies won't invest in factories if they think the demand is temporary. Because by the time they come online the demand will be gone.

I think it’s a logistics problem, so adding more supply via new factories (unless local) won’t resolve the problem if the goods can’t be shipped.

I would strongly suspect that corporate bulk orders are not currently affected, so Backblaze won't care until it gets much, much more dire.

CPU, GPU and Hard Drive.... guess we can have more dumb devices for now instead of everything IoT.

Wait until I unveil IoTCoin, it'll use PoIoT to verify the blockchain. Keep an eye out for my ICO!

/s (or is it?)

HNT Hotspot Miners are uncomfortably close.

My advice is to ignore anonymous new accounts making inflammatory and fear mongering statements.

Open Amazon and look up hard drive prices and availability.

This one is easy to fact check.

They appear to be normally priced and readily available.