Want to destroy bitcoin? I am going to tell you how.

For every person that advocates bitcoin it seem there is at least one person who will tell you that it is a colossal waste of electricity, effort and time as well as being some sort of scam.  If you are one of those people, this article is for you.

There is a badly kept secret with bitcoin, it is vulnerable to attack.  The attack itself is technically quite easy to execute and would result in the total paralysis of bitcoin.  Properly executed you would be in control of the entire bitcoin network and you would be able to bring the entire thing to a screaming halt destroying bitcoin once and for all (or at least as long as you can keep the attack going).

To mount this attack first I will need to explain a bit about how bitcoin actually works (sorry about that, I promise to try to not to get too technical).

The bitcoin network relies on the miners using their hardware and software to do what amount to brute force attacks to “crack/guess” the correct SHA256 (an encryption algorithm) to solve a block.

Each attempt to solve the block requires the miner to calculate a hash value for the block they call these attempts a hash, and the speed that they can make these attempts is called their hash rate.  When they calculate the correct hash they collect a reward (25 bitcoin at the moment) plus any transaction fee’s included in the block.

That block is then appended to the end of what is called the blockchain validating all of the transactions in that block and every previous block (and every transaction in those blocks).  This validation cements all transactions in place making them virtually impossible to reverse.

Attacking the miners directly is close to impossible, they do their work behind a Peer to Peer network designed to resist any sort of direct attack called “The Bitcoin Protocol”.  Peer to Peer (also called P2P) networks are notoriously hard to stop or even disrupt, the major Movie & Music Industries have tried to stamp out another P2P network you probably have heard of before called bittorrent with no success.  Even if governments were to outlaw mining (all 196ish of them) people say they will still do it (like the bootleggers did in the US prohibition years).

Attacking the transactions is also close to impossible because they are stored in everyone’s copy of the blockchain and crypto-graphically verified by the mining process, it’s a sort of open ledger that anyone can read and verify themselves.

So to bring down bitcoin you are going to have to beat them at their own game, fight fire with fire if you like, and use the mining and bitcoin protocol against them.

What you will need to do is build yourself your own mining operation just like the miners do but you are going to have to build one bigger and better than any of them, your mining operation is going to have to be able to equal everyone else’s hash rate plus a bit more (it is called the 51% attack by the way) and then you are going to start verifying blocks just like they do but you are not going to include any transactions in them!  In effect, bitcoin will continue, but no one can transfer their bitcoins to other users.

Lets do some quick calculations so you can get a shopping list together, we have all the numbers because the blockchain and network hashrate are all public information.  First we can see that the bitcoin network is capable of about 25 Thash/s (or 25,000,000,000,000 of those hash/guess’s per second) so we will have to do a bit better, lets say 26 Thash/s.

It would be good if we could use the new ASIC mining devices (Application Specific Integrated Circuit) were available, a single high-end ASIC device  is said to be capable of up to 1,500 Ghash (or 1,500,000,000,000 hash per second) for $29,899 USD.  18 of those devices would net you a respectable 27 Thash/sec for just $538,182 USD but unfortunately they are not available yet.

Even when they are available, according to the unofficial wait list there are already at least 24 high-end devices on pre-order as well as a plethora (always wanted to use that word) of other lower-end devices totaling a staggering 53.9745 Thash/sec meaning that the total available hash rate will soon jump to somewhere in the region of 75 or more Thash/sec.  This also has the effect of putting a time constraint on your 51% attack, it must be done soon, before the advent of the cheap ASIC devices in the next few months.

With the current high-end FPGA mining devices (Field Programmable Gate Array) capable of 26.4GHash/s (26,400,000,000 hash per second) you will need 985 devices coming to a total of $15,065,575 USD but alas there are not 985 of those in existence and they no longer seem to be in production (with the new ASIC’s close on the horizon I don’t imagine many people have been buying them).

That all taken into account it seems that you will have to use good old tried and tested GPU’s for your 51% attack.  A current high-end GPU overclocked to the max (ATI 7970) will set you back as little as $379.99 each for a still decent 828Mhash/s (828,000,000 hash per second).  Unfortunately you would need 32,698 of them for a total of $12,424,913 (AMD may give you a bulk discount though if they have that many).

Add to that you will also need the computers to run them, so assuming 3 ATI 7970’s per computer you will need 10,900 computers with high end main-boards and power supplies for around about $500 each totaling $5,450,000 USD.

You will also have to pay for power, a very good Industrial Rate Cost per kilowatt-hour is about $0.07 USD so your 9.177372 Megawatt (estimate) will set you back $15,417.98 USD a day, or $462,539.55 USD a month.

I am unsure of even where to start to estimate a place to put this all, internet connections or cooling cost but I am sure you can find something, perhaps an old nuclear reactor or something (Even better, inside a dormant volcano, you could have sharks with lasers on their heads).

So there you have it, for a small price of $20 million USD or so and an ongoing cost of about $1 million per month (you will have to keep buying equipment to keep pace after all as you need to continue the attack indefinitely) you could paralyze bitcoin.

Note: You could also use the above equipment to mine about $50,000 USD a day or $1.5 million a month.

Also Note: All dates and figures are an estimate as at December 2012.

Authors note: If anyone actually wants to attempt this, I want a job!  (I already have one, but I have always wanted to be an evil genius)

Neil Fincham, the MineForeman has over 20 years experience in the computer industry and runs the MineForeman mining operation for the co-op members.

Posted in Bitcoin, Editorial, Hardware, Mining, Sunday
11 comments on “Want to destroy bitcoin? I am going to tell you how.
  1. Not Required says:

    Step 1: Spend $20m
    Step 2: Make bitcoin unusable
    Step 3: Wait a few months
    Step 4: Miners start using ASICs
    Step 5: Bitcoin becomes usable again
    Step 6: ???
    Step 7: Loss

  2. So, if 25Thash/s will crack a SHA256 hash every ten minutes, how long would it take for 25Thash/s to crack one AES-256 cipher? I’m curious because I’d really like to know what it would take to open up the encrypted Wikileaks insurance documents that have been circulated.

    • The bitcoin blocks use a deliberately weakened sha256 has to make solving blocks actually possible we call this weakening the “Difficulty”. Currently the difficulty is set at 3438908 so that a block is solved every 10 minutes with the current hashrate. The greater the difficulty the longer it would take to solve the block.

      A full strength properly implemented SHA258 encryption would have the equivalent difficulty of 6,582,018,229,284,824,168,619,876,730,229,402,019,930,943,462,534,319,453,394,436,096 so at the current 25Thash/s the bitcoin network would take 3.64E+52 years.

      Suffice to say the universe would probably have ended by then assuming you never upgrade your technology. (all rough numbers)

      That said, you might get lucky and solve it in 10 seconds.

      • Seth says:

        Thank you very much for clearing this up for me. It’s always been a point of confusion on my part.

      • JulioHM says:

        Thanks! That’s exactly what I was thinking. If you double the hashing power of the whole network, the difficulty will increase (working as designed) and you would need much, much, much more than 17Thash/s to accomplish this. Seems to me that the cost of it is really impractical, even for a strong government like the US.

        I’m sure they have enough money for it… but it would leave a huge dent in their economy… and good luck explaining tax payers why they spent billions of dollars to stop an open technology from flourishing.

  3. cypherdoc says:

    lol. is this article supposed to be serious? don’t you think this would’ve been tried already almost 4 yr into this project? trying to buy that much equipment on the open market would drive prices for those components to the Moon and the price of Bitcoin would probably rise correspondingly. who taught you economics?

  4. What about using an existing mainframe… Are there public/privately held systems with this capability?

  5. ABC DEF says:

    This ‘secret’ is well-known theoretical flaw and is no secret at all. However, the most powerful available cannot match the current hashrate of the network (not even close), and this is going to get truer as bitcoin expands.

    And even if you could gather that much computing power, the protocol can always be altered in various ways to foil those kind of attacks. True, there is no central authority in bitcoin but there is a moral one (Gavin). If Gavin decides to fork the protocol most people will follow him.

1 Pings/Trackbacks for "Want to destroy bitcoin? I am going to tell you how."

%d bloggers like this: