10,000 Bitcoin Mining ASIC Chips Arrive in Switzerland – 3 TH/s | Hacker News Reader

10,000 Bitcoin Mining ASIC Chips Arrive in Switzerland – 3 TH/s

thegenesisblock.com

37 pointscgi_man53 comments

I'm fairly new to the Bitcoin world, but it seems to me that ultimately (in 10 years or so) one of two things will happen:

1) Bitcoin will gain large-scale adoption. As a result, the value of Bitcoins will greatly increase from what it is today.

2) Bitcoin will fail due to government regulations, better alternatives, or other unforeseen reasons. As a result, Bitcoins will become effectively worthless.

Am I oversimplifying so far? If I'm not, then it seems like investing substantially in Bitcoins (purchasing coins) now will either make me very rich in 10 years, or I'll lose my initial investment.

By the same logic, if I earn coins by mining now and hold them, I will either be rich in 10 years, or I'll lose my mining investments.

I'm not interested in diving into mining to earn a quick 10-20k while the market fluctuates, so as a long term strategy it sounds like investing is the more sensible and less time consuming option. Thoughts?

A lot of people agree with this thinking.

It's not the only possibility— e.g. Bitcoin could putter around for a long time basically where it is now, maybe acting as a very mild threat the keep the alternatives on their toes enough that Bitcoin never gets widespread adoption.

Right now it's possible to make a decent amount _today_ with mining as a small / hobby business, supplying coins to people who don't want to get into the mining stuff, without engaging in major speculation yourself.

I see this "puttering" as the most likely possibility, since if they do hit some level of adoption that starts to threaten sovereign currencies, the GP's option (2) will come into effect.

Maybe some non-democratic governments will become hostile to Bitcoin, but democratic ones will most likely not try to stop it because they eventually do what the people want.

Case in point: even the US Treasury Department's FinCEN bureau recognized enthusiastically that Bitcoin offers "potential" and "holds great promise" to US economy: http://www.fincen.gov/news_room/speech/pdf//20130613.pdf

Puttering would still be inflation-corrected.

Maybe. Puttering could still have a small positive or negative slope.

That sounds pretty much correct. I compare investing in Bitcoin to investing in an early stage startup.

Regarding mining, each time Bitcoin mining technology improves there's a mini gold rush. If you can get your hands on the technology before everyone else you can make quite a bit of money for a few months. But there's also the chance your preordered miner won't be profitable by the time you get it. This may be why hardware manufacturers prefer to sell hardware (i.e. pickaxes) rather than mine themselves.

Long term, once Bitcoin ASICs catch up to the state of the art Bitcoin mining will be very a very low margin business. Only the most efficient hardware located in places with cheap electricity will be profitable.

Such a waste of electronic computing power.

EDIT: what I meant by a waste of electronic computing power is that it does nothing for the world beyond pumping out more bitcoins. I guess it's more of my opinion on bitcoins in general than this particular artile. All the computational energy is not used to search for a cure for cancer, or aliens, or crack a code, or used to compute a deep neural net. I guess I would compare it to high-frequency trading---it makes people (a lot) of money, but doesn't deliver any net gain for society.

It's not a zero sum game.

It's not a "waste" of computing power at all, in my opinion.

- If BTC did not exist, this computing power would not exist. The demand for this power is in addition to other forms of demand. This isn't taking away power from any other field, it's simply new power being added for a new cause. - The technology derived from this "arms race" can benefit far more than this field.

Why wasn't the moonlanding a "waste" of money, talent and science? Because of the training it provided people and the tools and technologies they developed, etc.

Absolutely. I had a CS professor who lamented how much more time students spent playing games than studying, but acknowledged the fact that the video games industry is a gigantic stimulus to the computer hardware industry. Because games are always pushing the envelope of graphics and resource consumption, there is a reason for companies like Intel, Nvidia, ATI, AMD, etc. to produce better hardware. This hardware has led to benefits in the scientific computing realm, and may not have been developed without the bankrolling from gamers.

Can you briefly explain how the development of the custom ASIC's for BitCoin mining are of utility outside of their purpose built role? I don't understand that line of reasoning.

How many people are learning how to build ASIC's right now? Designing them, improving them. Scaling their production up. What kind of contributions are those teams making to the wider field of application specific circuitry?

What future benefits will these now-trained minds create with what they learned here?

What future technology will be built in some small way on the advances made in this field?

Future success isn't random, and it isn't driven by random new geniuses. It's built, person by person, experience by experience.

The men at NASA who got us to the moon learned skills that very few humans have been able to learn since. Their skills last a lifetime and have been brought to countless businesses and ventures.

I think the same thing here: the people, the skills and the technology are all beneficial to develop because we simply cannot know how they will be used in the future.

We can only provide a menu of options for the future, our future selves must build something new from that menu.

But if we do not provide a full menu, our future selves will not have what we need to succeed, and we may not even realize it.

> The technology derived from this "arms race" can benefit far more than this field.

Sure, I guess...

> Why wasn't the moonlanding a "waste" of money, talent and science?

The moonlanding was a major advance for mankind, independent of all the long-term benefits that have since arisen. You're comparing that to bitcoins?

You seem to not understand the potential benefits of widely adopted censorship-resistant (decentralized) currency. Try living in Argentina where the government is inflating your currency, while prohibiting you from moving your assets to USD. Try living in China where the government will seize your bank account if you are a business deemed too successful and if they deem they need a share of your profits. Etc.

That comment is pretty relevant, actually ... I (unfortunately) know actual human beings who feel that the moon landing, along with all current mars missions, were and will continue to be a waste of time and money. I'm completely baffled by it, but the fact remains that there are people that consider it no more valuable than bitcoins themselves.

What about the electronic power and human work that goes to operating banks? By that logic, the alternatives to Bitcoin are much more wasteful.

Edit: What I was trying to say, that maybe wasn't quite clear, is that Bitcoin mining isn't wasteful because it helps protect the integrity of the Bitcoin network. Other methods for holding and transferring money requires trusted third parties, who also "waste" (probably much more) energy, time and resources to operate.

Thats a fair point but why should I believe in bitcoin not another crypto ledger? The idea is fine, and the idea of a validatable record is great, but why should I support this one?

That's a complex topic with no clear answer. Philosophically, Bitcoin leaves much to be desired. But Bitcoin is the only realistic choice. The others (Litecoin, etc) just don't have enough momentum to be viable right now. Network effects make Bitcoin the only cryptocurrency with any short-term value.

However, regarding the philosophy: Bitcoin is a deflationary currency. That's a remarkable choice, because the prevailing economic wisdom is that inflation is a good thing because it keeps the money flowing rather than being hoarded. So most people probably wouldn't support Bitcoin's economic theory if given the choice.

It will be interesting to see how it turns out.

That could be said in the time of CPU, GPU or FPGA mining, but since these ASICs do nothing but Bitcoin mining, I think computing power is only wasted when they aren't running. In other words, I don't think they qualify as "computing power" for anything but Bitcoin mining.

But even when they're running all they're doing is pointless "mining bitcoin". And these ASICs are worse because they can't be re-purposed to do anything else.

All that computing could have gone to Folding@home and done some actual science. Instead it's just burning power and creating heat.

The same's true of mining equipment used to pull worthless diamonds out of the earth and the pollution it creates. The "problem" is the pointless mining generates valuable stuff in addition to waste heat.

...and keeping Bitcoin working, and possibly enriching the ASIC's owners. So it's not just burning power and creating heat.

> what I meant by a waste of electronic computing power is that it does nothing for the world beyond pumping out more bitcoins.

This is a deep, though common, misunderstanding of what mining is for. The purpose of mining is to achieve consensus and protect transactions against reversal. The dissemination of new coins is a wisely-aligned side effect.

The computation here is computation spend making the history of Bitcoin secure and permanent. It's not a waste.

All the work effectively makes bitcoin more and more secure, so it isn't waste at all. Potential 51% attacker has less chances to succeed the more miners there are.

Conversely, if these fancy chips become the standard mining kit, they will push regular hardware out even more. This means if you want to do a 51% attack there are much fewer individual people you need to compromise.

A 51% attack isn't based on the number of people, but based on computing power. If 30 people control 60%,then they control everything.

The more miners, the bigger the double spend required to compensate attack cost, the bigger the chances of it to be noticed.

I believe bitcoin is at the point where it is a financial loss to even try to compromise the network, even in case of success.

so, just like in the gold rush of 1849, those getting rich are those selling equipment to the miners.

It's more like "there ARE people getting rich selling equipment". There are plenty of miners getting rich.

This isn't manufacturing money out of nothing. This isn't generating physical materials with tangible value.

The only money being generated here is simply being vacuumed out of the pockets of investors losing money on trades or exchanges.

In that sense it's not unlike online poker.

If "generating physical materials with tangible value" is your benchmark, what are your feelings about Internet companies like Google and Facebook? They make very little in terms of physical goods.

Recall that Bitcoin allows people to send money around the world in an extremely simple fashion, and at almost zero cost. Banks typically charge $15-$20 for international wire transfers on both ends of the transaction. I grant you that there a huge amount of speculation, but surely the ability to transmit value around the world, securely and very cheaply, is worth something?

The thing is, Google and Facebook both produce something with actual value, be it content, or databases, or user data, or people's photos with their intrinsic value. Bitcoin is literally just numbers. It's not even pieces of paper. That's why it's better compared to virtual poker chips.

Also, the transaction fees for Bitcoin are surprisingly steep, especially if you want to convert it back to cash. You'll easily pay $30 per transaction for $1000 worth of BTC just to cash out.

Or the US dollar (or any fiat currency).

Plenty of people found gold during the gold rush. That does not mean it is even remotely representative of the norm.

Am I missing jargon or are they saying terahertz per second?

It's terahash where hash means double sha256 that is used in the bitcoin proof-of-work protocol

i think its tera-hashes per second

So... I agree that Bitcoin has some advantages, but it's disadvantages are quite high as well.

It seems to me that all the money going into Bitcoin mining equipment, and all the power being used to run it is essentially an indirect tax on Bitcoin usage... I wonder what the total value of Bitcoin transactions in USD compared to the cost of electricity to keep the block chain up is.

The true value that Bitcoin mining provides must be some low percentage of the total Bitcoin transaction volume... or am I missing something?

It is impossible for Bitcoin transactions to take place without miners. BTC miners provide cryptographic integrity, proving that transactions took place at a certain time between certain people.

Bitcoin is purely held up by the community of miners. The integrity of BTC becomes greater and greater the more trusted miners enter the system.

Couldn't this be achieved in some sort of other way? One that does not require hundreds of thousands of dollars worth of equipment and electricity? Couldn't a regular PKI message digest on the transaction list work?

I suppose there's something I'm not getting about this.

Bitcoin is a voting-based protocol. The "vote" with 51% of the computational power of the system wins.

By performing very very hard math problems with every PKI digest (ie, the "proof of work"), with a "difficulty level" such that it'd take the entire BTC network approximately 10 minutes to find a solution... then it becomes extremely extremely difficult for someone to enter the BTC system and screw up the public ledger.

You see, BTC is a distributed database, cryptographically signed by the BTC miners every 10 minutes. The difficulty of _performing_ the signature is automatically adjusted to take a ridiculous amount of time.

What happens when all 21 million bitcoins are mined? How will transactions be archived if no one is putting in the effort to mine bitcoins?

Transaction fees are expected to replace the block reward by the time that happens.

Which is year 2140 or so

The supply of "subsidy"— newly created coins— declines exponentially over time. So a long time before all 21m is mined, sometime around the year 2140 the amount of new coins created this way will be insignificant.

Bitcoin transactions can optionally provide transaction fees which are paid to the first miner to include their transaction. Blocks have a limited size in order to prevent the biggest nodes from centralizing the system by producing more load than smaller ones can handle, so there will be competition for access to that limited space. Miners can then be funded collecting these fees.

transaction fees

This basically will signal an end to the typical BTC pools. At .1 BTC (~$10) per chip, it is now possible to beat a $400 AMD 7970 with just $20 worth of chips.

This doesn't change the pools at all (pools don't change if everyone uses ASIC). This does probably signal the end of GPU mining though.

Fuck bitcoin, if custom chips are that much cheaper the GPU power in cracking hashes then most key streching algorithms are worthless.

Apparently people downvote me. I don't hate bitcoin, but if the relative power is true then this is so much bigger than bitcoin.

Lol. You made a good comment, I dunno why you got downvoted.

One thing to note: key stretching is a good idea nonetheless. hashing ASICs will be millions of times slower _regardless_. They may be hundreds of times faster than a computer, and hundreds of times cheaper... but the effect doesn't get you what a good Bcrypt-difficulty 12 gets you (ie: key-stretching with ~4million hashes)

What is this in current $/hr?

If my off-the-cuff calculations are right, then a little under $10/hr for the whole batch of 10,000 chips that just arrived(i.e., each chip will generate about $0.01/hr), which seems pretty low until you realize that each chip costs only about $10. Break even point is about 40 days, right? Not bad at all. Unfortunately, difficulty will probably spike, so these rates are not sustainable. And who knows what the price will do.

Unless you're a very shrewd businessman and want to do this full-time, you're far better off just buying Bitcoins if you're interested in speculating on them. Just remember that Bitcoin prices will inevitably plummet, go sky high a few months, and then re-plummet. It's just what Bitcoins do.

http://i.imgur.com/zAMkNCV.gif

Reply on news.ycombinator.com