Another Bitcoin death. Now ASIC

Already many times on Habré the death of the only "non-state" Bitcoin currency was foretold, someone said that it was not a tenant by itself , someone that it was being hacked by some hackers , and some even gave practical advice on handwritten annihilation.

And yet, despite all the problems, threats of hackers and other troubles, now the Bitcoin system is alive and well, but only because they have not stopped believing in it.

Now a new threat has appeared - a threat already from those who, despite all of the above, continued to dig, buy, sell and use Bitcoin in every way. And the name of this plague is ASIC.

A bit of history

At a time when the Bitcoin system just appeared, the only way to “earn” coins was to dig these very coins (mining) through the CPU. Mining, in a simplified form, was a solution to a cryptographic task, according to a specific, strictly defined algorithm. The reward was supposed to solve this problem - one Bitcoin block, which contained 50 coins. Through each specific number of blocks mined, the complexity of the task that had to be solved increased - this is such a simple and uncomplicated way to avoid inflation of a given electronic currency. In principle, these basic rules of the game have not undergone any changes since the shaggy times, but the mining process has undergone significant changes.

PC processors in their current state were not powerful enough to solve the problem of finding Bitcoin blocks. Although the word “powerful” is somewhat inappropriate here, it would be more correct to say that they are not able to effectively solve such problems. And the complexity of the tasks does not stand still - each time it is necessary to sort out more and more options in order to find the right solution. And instead of a couple of hours, for which the block was usually located, it was necessary to spend a day, or even two, to find a solution. And here the miners (as those who “mine” this currency call themselves) turned their eyes to the most powerful “number crushers” that could be obtained by a simple user - video cards.

The increase in speed was simply amazing - is it a joke that the current Intel “top” produces a speed of about 40 Mhash / s (this is the designation of the speed of solving problems adopted on the Bitcoin network), and the old Radeon 5770 video card was already issuing 200 Mhash / s. The difference, as they say, is on the face.

With the beginning of using the GPU, the complexity of tasks began to grow exponentially, but despite this, in those days, the money spent on buying a video card fought off in just a week or two, and after that, mining began to make a profit. But over time, the complexity of the tasks again began to approach the threshold of "unprofitability." In a simplified form, the cost of production is considered the cost of electricity consumed by the video card, plus the depreciation of this video card. And while this figure was less than the cost of mined coins, we can say that mining is “profitable”. And in order to cross the threshold of unprofitability, which came at the time of unnecessarily complicating the task, a new solution was found, which was called the "farm".

A farm is the nth number of video cards connected to one computer, and on one motherboard they manage to connect 6 video cards, even connecting them via PCIx1. It looks something like this:

In contrast, ordinary users who do not want to buy a bunch of expensive video cards, and bother with their connection and setup, began to merge into "pools". About pools on Habré there is a wonderful article , I recommend that you read it.

Well, now let's talk about the current state of affairs.

The present

Currently, mining in the pool, in fact, is the only way to somehow make money on the Bitcoin network. Moreover, even the owners of large farms are mining in a pool, solo mining is dying out as a class. A person with a single video card, if, of course, he is not the owner of the current "top" Radeon 7970, except how to pay the electricity bill, mining will not help. The complexity and number of earners is growing, profitability is falling.

In addition to pools, various kinds of “financial pyramids” based on coins began to appear, some provide their computing power for a certain deposit, some work like the well-known MMM - bring coins, bring friends, get 10% more.

Of course, all this cannot save the project, only to prolong the torment a little. A new mining method was required that, due to the increased complexity, would provide a reasonable mining speed.

And of course, there were people who invented such a method - it turned out to be Butterfly Labs engineers . They created specialized expansion cards that only knew how to dig coins, but they did it well. Even the first generations, at a penny of power consumption, gave out speed comparable with the best representatives of the GPU. Of course, this did not greatly affect the overall picture, after all, the GPU device is more functional and at approximately the same cost, users preferred video cards.

But the time has come for a new development of the "Butterfly Labarotaria" - ASIC.


ASIC is a family of specialized processors. who in life can do nothing but solve the problems of Bitcoin. They use the SHA-256 hashing algorithm (a special case of the SHA-2 algorithm ), thanks to this the mining performance of these babies is simply fantastic - the youngest model, of the BFL BitForce Jalapeno products developed on the basis of ASIC (test sample), it produces about 3500 Mhash / s . For comparison, the current top Radeon 7970, with overwhelming overclocking, produces 800 Mhash / s. And again, the "difference on the face."

It is not known for certain what is used as the CPU on these boards - test samples were not given to anyone, and delivery to customers will begin only in November. But the Bitcoin community suggests that this is a modified Spartan-6 chip, which was previously used in the development of the "Butterfly Company".

What will happen

And this is what happens - bitcoin will die like free currency, and will become just “candy wrappers” suitable for paying for services on underground sites like SilkRoad.

Surely you ask "Why?". Let's estimate the development of events. For this, I will give a few numbers.

TitlePerformance in Mhash / sPower consumption wattsCost $
Radeon 7970825214420
BFL BitForce Little Single30,00040649
BFL BitForce Single 'SC'60,000601,299
BFL BitForce Mini Rig 'SC'1,500,0001,50029,899

I think from the table it becomes clear that in the current alignment of forces, the GPUs will be out of work. Moreover, when compiling the table, I even cunning a little - the values ​​for Radeon were taken the best, of those that miners were able to achieve. Its average productivity will be only 600 Mhash / s. But believe me, this does not play any role in the current state of affairs.

Now let's arm ourselves with a calculator and count a little. It turns out that in order to gain a farm, with a performance commensurate with the junior model BFL BitForce Little Single, you need to buy 38 video cards. And this is $ 15,960! Which, in turn, is almost 25 times the cost of this specialized digger! About how much electricity will be consumed and how the farm from 38 Radeon will warm up, I better keep silent. I can only say that you will not have to pay for heating in the apartment.

From an economic point of view, the conclusion is obvious - GPUs are now useless. After all, even digging in the pool, you get a share proportional to your contribution to production. And there’s no more talk of digging in solo mode, it will take a year, if not more, before you find the block.

Now we abstract from the economy and look from the other side. Digging coins for me has never been a way to make a profit, but an opportunity to earn a couple of hundred rubles. In order to earn them, I only had to leave the computer turned on for a while while I was not at home, and it would be foolish not to use it. Fortunately, I change video cards over the course of a generation and always take the top, which is why my production speed is not bad.

What now? And now, even mining coins 24 \ 7, for a month I will at best get 50 rubles. The complexity grows in proportion to the speed of production of new blocks, and with the advent of ASICs, it will simply take off. Of course, the idea of ​​getting 50 rubles for hell-knows how much time, when fully loaded, the GPU is simply stupid. It turns out that in order to leave my production rate at the current level, I need to purchase one of these "diggers". It would seem like a good way out, and it costs not much more than a video card, and digging performance is much higher, but there is one thing - it’s not universal. This is the main reason why Bitcoin will leave the masses, no one will want to buy one more thing for $ 650 “in addition” to a good video card, just to dig coins. After all, this thing can do nothing but this, I can’t watch a movie on it, connect 2 more monitors to it, in the end I can’t even play new toys! The Bitcoin network will lose the main thing that gives it life so far - mass.

Despite all the problems, millions of people dig, dig, dig coins every day. Then they spend them, exchange them for another currency or invest in the "pyramids". Let the turnover of such "diggers" is many times less than those who dig on huge farms, it is these simple people who provide trustto bitcoin currency. But without faith in them, these coins will only become abstract "wrappers" for exchange. Like stickers or gum liners. You will never go to the exchange and try to exchange your collection of Love Is earbuds for rubles, although if there are those who sell something only for earbuds, maybe you will find someone who will exchange them for you for money. But the market does not consist of 2 people. There should be millions of them, those who buy, sell, produce, change, store ... But there will be only a few in Bitcoin.

When the largest Bitcoin exchanges conduct 10 transactions per day, they will become unprofitable. Moreover, how will it be possible to establish the exchange rate that only a small part of people will own? Here we get the problem as with the pictures - how many actuallyis Mona Lisa worth it? Priceless? But it seems to me that she costs exactly as much as they are willing to give for her. So it is here, each coin costs exactly as much as money, people are willing to pay for possession of it. Only a picture, unlike a coin, gives pleasure, pleases the eye, and just +500 pathos, because there is no one else who has the original Mona Lisa. A coin does not give any of the above, but as a currency it should give the right to exchange it for something from the real world. Even if it will be drugs, weapons and pornography - but it is these things that give the right to call Bitcoin currency, and these very things show us how much one coin costs .

Of course, with the advent of ASIC SilkRoad and other "hot" places where you can buy something for Bitcoin will not disappear, but most exchanges will disappear - the amount of coins that are now on hand will depreciate, thanks to the huge production speeds that ASIC will give us, and getting new coins will be very, very difficult. At such a moment, the seller on SilkRoad will be free to display his goods for any value that he considers acceptable, because in the Bitcoin world there will be an analogue of our default when the price of goods changed every hour. No one will know how much their coins are worth now and how much they will cost in just a couple of minutes.


After reading this, you probably thought: “after a month or two, the collapse in the value of coins will stop and normalize, it has already happened more than once,” and you will be right, but again, the main problem is not that your coin holdings will turn into nothing, but that you can’t get more coins in any other way than to buy them. After all, getting them in the usual way, having dug up in a couple of days, will become impossible. And to buy a stray, even for $ 600, just for the dubious opportunity to buy something on SilkRoad or to receive, even 1000 rubles in the first two or three months (and then again the growth in complexity, again the problems with speed ...), it will be just stupid.

Everything written in this article is only my vision of the situation, and is not "true in the last resort."

Then let me take my leave, thanks for your attention and have a nice day!

Also popular now: