Cryptocurrency miners are renting Boeing 747s to ship graphics cards
pcgamer.com
pcgamer.com
Using a random card Amazon gives us a box size of 13.4 x 9.5 x 3.1 about 400 cubic inches. Say http://www.cargolux.com/our-expertise/cargo-equipment/aircra... we have 21 347 ft^3 that is some 36 887 616 cubic inches which means some 92 000 such boxes give or take.
There are inefficiencies etc so let's say it's 30 000 cards. Say at 300 USD each. That's a cool ten million bucks. The plane likely costs less than half a mil. It's not a biggie.
But do they really? https://motherboard.vice.com/en_us/article/zmemza/cryptocurr... says "Enigma started with an installation of about 700 GPUs and now consists of several larger buildings," Schindler told me in an email. "Obviously at this size the operation draws attention from the hardware manufacturers directly and we have agreements with them." Compared to that I have some doubts they need to fly thirty thousand cards in. Maybe they rented a pallet in a 747F...
I don't have any insight into what they are currently up to because that scaling apparently involved hiring a eastern European engineer on the cheap to masquerade as a customer to try and obtain confidential hardware design details from me so he could rip off my product for them.
Sounds like an interesting story here. How did you out him/her?
i imagine it to be somewhere between Hackers and Departed
PS: given we currently have an emoji movie, had an facebook movie. A bitcoin movie is only a question of time - too many good stories here.
If there are elements of Hackers in it, tensorflow and kubernetes is killed off by Apache's many offerings. The Gibson will be moved to the SMACK stack. Google is killed off because of a lack of any real customer service. Apache controls the ledger and the hero hackers will convince China, Russia and the USA to take all of their super computers to win back the ledger.
The business operates on a model where the customer pays a minimal fee, $100, to purchase a prototype of a new design but I retained the rights to the designs and did all the design, development, and testing on my own dime and then amortize those costs over the life of the product.
So one day they come to me and say they want to start manufacturing at a factory they've made a deal with in China and ask me to give them the design files free of charge. I told them that I wasn't going to just give them the companies IP for free because a lot of work went into developing it etc... but that we could discuss what an appropriate price was if they really were dead set on doing their own manufacturing.
There was some back and forth but their position was that they didn't think there was anything special about the product and the most they were willing to pay for access to the design files and a license to manufacture it in unlimited quantities was $100 per design. I pointed out a number of things that we did the most important of which was making sure the design and materials could safely handle the high current loads the boards had to withstand, etc...
Shortly after I refused their $100 offer as insultingly low I get an email from some guy purporting to be from eastern Europe and saying he was interested in purchasing 500 units. Before he would go through with the purchase he needed to confirm some technical details of the board design because he was "concerned his environment was very hot and dirty". He then went on to ask a list of questions including things like what are the number and thickness of the copper layers in the PCB. These were the exact reasons I had discussed with the Genesis guys as to why our designs had value. At that point I was 99% sure of what was going on. So I responded that we didn't give out those specifics but that we could send him a prototype to test and he never responded.
To satisfy my own curiosity a little while later I used a personal email address to send the guy a message saying that a friend had given me his name and asking if he still did hardware design because I needed some design work done for crypto-currency related projects. A day later I got a response that removed any remaining doubt about what was going on.
It used to be called 'documentation"...
I never know if I'm being shilled or not anymore...
Here's the reality--
-- Ethereum is going to PoS soon. Even if it doesn't fully switchover it will decline in profitability for miners quite a bit.
-- Ethereum is down %50 in the past couple weeks. The GPU boom is over (for now) and supply should be loosening.
-- There are just not that many coins out there that are GPU mineable to support really massive operations, like a full 747 of GPUs. BTC is $45B and ETH is $18B market cap. As far as GPU mineable coins looking at the top hundred on CoinMarketCap I see about $2-$3B market cap for them all combined. You aren't going to take the hash power of securing ETH at $18B and dump it into ETC and ZEC without thinking twice, and if you do you aren't going to be shipping 27,000 GPUS in at a time 2 months before you have to jump from your big pond into those tiny ponds.
-- And if there were, the labor to bring that many machines up, even if you already have the physical and electrical capacity is weeks of work, unless you're employing a very large labor force which you are going to turn around and fire right after they've built your very cost intensive and valuable and hard to secure warehouse full of miners.
Reality is, the industry may be buying one 747 worth over the next several months but they are all getting them a few at at time and constrained by electrical and physical capacity and labor availability, and parts of everything else.
At even low scale these are non-trivial issues, when you start building 10 rigs at a time it is at least 10 times harder, 10 times more likely a part is out, etc etc.
Also you seemed to assume this airplane-load of GPUs is meant for Ethereum, but Genesis Mining (the miner who rented a 747) already mine ZEC and other alts.
Finally, I would wager there are probably more (at least as many) GPUs mining all other coins combined than GPUs mining Ethereum (2 millions or so).
But you are right there's a huge number of GPUs out there. My main point is there is going to be a big shift to the others from ethereum when ethereum becomes less profitable...
A ticketed courier could deliver a large number of cards just as quick, if not quicker.
Or just because it sounds good in an article?
Exactly. I bet if you took the cargo capacity (I'm guessing that with any packaging to speak of, volume would be the limiting factor rather than weight) of an air cargo configured 747, months worth of Genesis' card consumption would easily fit in one flight.That one hyperbole quote generated at least a half-dozen different articles in this media genre, most of which have been posted multiple times on HN over the past 4 days.
Not for long. https://techcrunch.com/2017/07/25/sec-regulators-are-coming-...
Could you elaborate?
Bitcoin could hypothetically be replaced by Dogecoin and bitcoin would have no value.
A cryptotoken is still a cryptotoken as long as the network is propagating tx & blocks. The whole point of them is that they continue to function regardless of if the "developer" is around.
You can hold and you can dollar cost average, and if you weren't buying with the intention of holding for 5 or more years you weren't really investing you were speculating and if you're speculating you should be hedged or have a strategy that can handle a dynamic market.
I know people who were early in bitcoin and they are still HODLing and using their BTC wealth to invest in other things like ETH and startups.
I think I understand what you mean, but I also think that you underestimate the impact that the introduction of the tulip to Europe had on the economy. If you browse through the Wikipedia page, I think you will find it interesting: https://en.wikipedia.org/wiki/Tulip_mania
Also interesting is the effect that it still has on the economy. Some of the big flower monopolies that still exist today (nearly 400 years later!) were created by people who speculated on tulip bulbs and diversified quickly enough (mostly by using profits to buy up land).
We tend to think about things from our own perspectives, but when you are looking at that time period, imports of agricultural products to the wealthy is what ruled the world. Just like they were able to saturate their market, so will we with abstract financial products. Those who are able to diversify quickly enough will probably be the next big winners. Those who keep buying graphic cards will probably be the next big losers.
Bitcoin has significantly higher transaction fees than paypal and a much less mature, much harder to use API.
However, simply holding bitcoin as a hedge against inflation and central bank currency risk is a legitimate use.
One of the big rallies in bitcoin coincided very precisely with the Cyprus collapse where the government simply confiscated a percentage of everyone's money in the banks.. and in places like Venezuela bitcoin is literally a life saver.
So remittances may be the biggest market, but hedging against incompetent government is not an illegitimate use of it... and I would separate that from "speculation" as a use case.
Speculation tends to get a bad rap, but remember, speculators sell when prices are high and buy when prices are low-- they tend to moderate prices.
I heard that last time before the crash. Kind of a weird statement to make now that it's at $2800.
The fact that they seem to be pretending that this use case does not exist and stubbornly declare themselves to be exclusively a gaming company is baffling to me.
Brief research suggests that the entirety of Etherium mining is roughly equivalent to 1 quarter of AMDs GPU sales. It seems gaming is the much bigger chunk of demand (Nvidia sells 3x what AMD sells, so annual gaming sales are roughly 10x the total Etherium spend).
a REIT is just a special case of an IT eg "tritax big box" vs a RIT Capital Partners or Wittan
It would also be wasted effort on AMD's part as the cards are selling like crazy anyway and once the cryptoboom stops there's still plenty of gamers left to buy them.
If they made mining-specific cards with no video output, then they would be useless on the secondhand market
* Overproduction for the sake of mining can leave you with huge oversupply of cards, which is very costly. * GPU miners can crowd out the enthusiast market, which is terrible in the long term: * gamers have brand loyalty and buy peripherals (like FreeSync monitors) and miners don't. * if gamers can't buy GPUs at affordable prices when they're building their new PC, then (some companies perceive that) they might just buy consoles instead, which shrinks the entire market. This is more a problem for the "budget" market, obviously. * Enthusiasts recommend products to non-tech-savvy consumers, which again, means that if temporary mining sales crowd out non-mining sales, they lose money in the long term.
The problem is their core audience (gamers) gets cranky when they can't buy cards at regular prices due to the huge crypto demand. It negatively affects their image and they would rather have their fanboys than the disloyal miners who would happily switch to nvidia if it meant a greater ROI.
https://news.ycombinator.com/item?id=14889357
Apparently they are going to bundle most of them with games and hardware discounts initially, to discourage buy ups by the miners.
http://www.anandtech.com/show/11680/radeon-rx-vega-unveiled-...
The RX Vega 56 part has 50% higher SP compute per watt than the RX 580 and 570 (50 SP GFLOP/watt), and is only $399 MSRP.
Single Point calculations per watt or per second are not really the relevant measure for miners, it's all about memory bandwidth.
We will see how these stack up.
https://www.techpowerup.com/reviews/AMD/Vega_Microarchitectu...
But when it comes to money, greed and self interest always triumph and there will be dozens stretching logic and apologism as far as it can go.
'Proof of work' has to the most hare-brained, wasteful, anti-environment scheme ever thought of. Responsible people should not just be wasting electricity in this manner.
https://motherboard.vice.com/en_us/article/ypkp3y/bitcoin-is...
Further, the cost per transaction in bitcoin a year from now will be about 1,000,000 times lower due to Lightening Network.
I know there's a political desire to demonize bitcoin (this is vice, not exactly an objective source) and that ignoring one side of the equation to try and make bitcoin out to be an environmental evil seems the best approach you can get but its' absurd on its face.
While I like the basic idea behind Bitcoin, I really don't like this development of "Let's invent a couple of dozen more of those!". It also shows the ugly face of how what we as human societies actually value. While I haven't compared any numbers, I'm pretty sure Folding@home or Seti@home can only dream of having the kind of computing power available to them that some of these cryptocurrencies manage to mobilize.
You want my cycles to find a cure for cancer? Nah, I'll keep my cycles to myself.
Using my cycles to make the imaginiary profits? Sweet, let's buy warehouses full of hardware!
People have attempted to make scientific computing good for proof of work-- GRIDCoin is an example. But it doesn't work for actually securing the network, due to the nature of how blockchains work.
You're free to work or mine or do whatever you want and then donate that money to scientific research.
But what makes you think you know better how I should spend my money? You don't even know how I'm spending my money.
I should have clarified that more: As I said before I have no issue with the basic concept behind cryptocurrencies, Bitcoin is a great thing and will hopefully do even greater things in the future.
And of course, everybody is to free to do with their time and resources as they please, never suggested otherwise.
What I have an issue with is how new currencies seem to be popping up like mushrooms because they've become the current "thing" to do if you want to make a quick buck. From the outside, it just looks like a bunch of pump and dump schemes, with no purpose besides offering a couple of early adopters the chance to make temporary huge profits, comparable to a digital Wall Street.
Originally I imagined BC as the solution to all the exchange rate shenanigans and whatnot, have one global and digital currency for everybody. Instead, we are now forking currencies like software, I'm not sure I like the possible long-term ramifications of that. Do we really want a future with hundreds, if not thousands of different, and competing, cryptocurrencies? Is such a future even viable?
It's not like we haven't learned these lessons before in the real world, we have, plenty of times. That's why it's so surprising how keen we seem to be on repeating them in a digital format for the sole sake of having yet another speculative vehicle from which we can deduct some imaginary "economic growth".
It's really hard to tell what category something is going to fall into... but the thing is, bitcoin has had massive returns and if you can produce a really good innovation, the impact on society and consequently the returns could be equally massive.
So yes there is a lot of speculative money out there, but the only way to tell the good from the crap is to do your due diligence.
For me, even if the good is %1 (and I think it's a bit less than that right now) it's ok.
That is a loaded statement. If you ask a monk whatever you are doing in life has no real purpose.
A centralised service like HN on the other hand could be designed to scale linearly with the amount of hardware resources.
Edit: Now that I think about it I wouldn't be surprised if someone makes a cryptocurrency based around "digging holes" keynesian style. I've even found a name for it: captchacoin.
Hell, I don't even expect you to agree with me. I'm just posting to speak my point of view and amuse myself. This conversation is pretty much the most wasteful thing we are going to do today.
Real life is less utilitarian than many people wish. Even if our conversation is more efficient, it generates not one bit of value - at least financially. It does amuse me, but I'm not really sure that's very valuable.
That's valuable. It lowers the demand for jetskis and cocaine, at a fraction of the price. Also, there's not much point in life if being entertained isn't valuable.
You are free to design a better alternative. If you do it, you will be rich.
It has no value to the wider economy or world but to us, and even though it has value to my friends it's a terrible, wasteful and unethical way to generate value. I am not generating value, just wasting trees.
You just can't create money or value out of nothing. Its linked to utility, markets, labour and the wider economy and belies the fundamental confusion in the crypto world about money. But intoxicated by the prospect, like alchemists and other charlatans before them the crypto community continue to fantasize about gaining something from nothing.
That wood builds your house, provides heat, provides jobs, makes dowels, and makes paper - depending on what is harvested. And, it will be able to do so in perpetuity. The process is called TSI and is even able to increase habitats for the animals - they are delicious.
I'm not sure we can have a productive conversation.
This is what current crypto currencies are doing. How long before people notice you are wasting wood for no purpose and put a stop to your activities?
Externalities* apply here. Global warming is one of them. Increased vulnerability to desertification is another effect.
Otherwise, we could just fill the Sahara with trees in a decade. It is not going to happen.
Would your criticism extent to gold?
You're argument barely makes a blip. Perhaps if you should choose not to browse reddit or hacker news at least 3 times per week to save the environment.
I think this is a general shortcoming of proof-of-work. I wonder if this could be fixed with a different scheme, e.g. reputation-based, or proof-of-(something socially desirable and not centralizable)?
But you can still "take part" by using it for what it actually is - a currency.
I don't know anyone who would want to use BC for anything other than speculation. Speculation that undermines the stability of the currency, which isn't conducive to use as currency.
There are proper cryptocurrencies that don't require mining but they don't seem to get nearly as much attention.
- Turning my "being good with computers" / learning about it early on the internet into money
- Buying illegal stuff on the darknet (which I didn't do yet)
- Speculating / Gambling
- Transferring wealth across borders while avoiding high conversion fees
While these are typical money functions, they are all fringe activities. I don't see myself ever using bitcoin as a replacement for Euros - that is for receiving my salary, for buying things (that I can get legally), for investing or saving.
One reason is I don't share the crypto-idealists idea of an anarcho-capitalist utopia...
AMD or NVidia could 1up them by air shipping, which is very low cost considering the value per kg of the shipment.
But AMD cards use less power with integer performance.
AMD is generally more popular among hobbyist miners, I think, because early on they had a performance advantage in Ethereum and hobbyists tend to follow each other like a herd because they have limited time to do experimentation and exploration (and this is not a criticism of them, they have day jobs and it takes a lot of effort to even get miners going, let alone optimize them across multiple parameters.)
[1]: https://motherboard.vice.com/en_us/article/59zdjq/an-idiots-...
If you are the manufacturer, why not let the customer take the risk? You are planning and spending years ahead, why gamble more than you need on cryptocurrency price?
On the other hand they would have a self interest to preserve the system, if trust in cryptocurrencies collapses then the big miners would have a lot to loose.
But that' d knee cap the value of the currency. Which in turn would nuke their profits and render their capital outlays useless.
It's really expensive attack to get back the $5 you paid for Starbucks.
So you would want to do it with a multi-million dollar transaction.
However, the thing is, anyone you pay a multi-million dollar transaction to is going to let it clear for more than just one block... probably for many blocks...before handing you the goods.
Because everybody is aware of this risk.
Chains reorganize anyway, just naturally when sometimes two blocks are found within the propagation delay.
Trying to run ahead is very hard.
Wallets like Electrum show one confirmation in red, and several in yellow and only after the chain has gotten extremely expensive to unwind will it show them in green.
I imagine the factories are where they are in part because land and power is cheap?
This is no different from the tulip mania of the 17th century. Yes tulips still have value today and cost something after all but not the absurd amount of money they once fetched.
seems like the average person barely breaks even after paying the home electric bill.
[0]: https://qz.com/953614/california-produced-so-much-power-from...