With bitcoin, the popular cryptocurrency that has been making headlines of late, governments are scrambling to determine how to regulate it before it’s too late and investors are entering the field by droves (even my mom asked me what bitcoins are). In this regard, there is little we can say that hasn’t been said before.
We are more interested in the technical details of bitcoin, the fine print if you will. In this post we are going to delve a bit deeper into the network dynamics and argue that if the current consolidation trends continue, the possibility of an attack by an “evil/dishonest player” will grow by the day.
It is a well-known fact that if an attacker could achieve the feat of single majority (compute more than 50% of the hashes) he will be able to take over the block creation. This attack is known as the 51% attack.
The bitcoin wiki explains it pretty succinctly :
“An attacker that controls more than 50% of the network’s computing power can, for the time that she is in control, exclude and modify the ordering of transactions.
This would allow her to:
- Reverse transactions that she sends while she’s in control. This has the potential to double-spend bitcoins that previously had already been used for transactions in the block chain.
- Prevent some or all transactions from obtaining any confirmations
- Prevent some or all other miners from mining any valid blocks”
While the actions such an attacker might execute do not seem to be very significant, an attacker with vested interest in not allowing Bitcoin to succeed might have the motivation to go even further. The damage that a successful attack could have on Bitcoin’s credibility is so high that we have seen examples of miners voluntarily dropping out of pools that could threaten Bitcoin’s stability, as occurred in the Ghash.io pool.
It is a widespread assumption that the cost of mounting such an attack at the current difficulty rates is high enough: not even large-scale government attacks can be mounted. We argue that this is not the case.
Given the computing power necessary to mine bitcoins, trying to do it with today’s CPUs or GPUs is child’s play. ASIC hardware is the only realistic alternative given the computer power of the network currently available. To put this in perspective: at the time of writing, the fastest AMD Radeon 7970 x 3 GPUs can compute an average of 2GH/s (gigahashes per second). On the market, you can find medium range ASIC capable of performing up to 50GH/s and the fastest to date is in the ballpark of 3TH/s (terahashes per second).
The raw power of the ASICs has changed the landscape of mining bitcoins forever, and there is no going back. This trend will become far more entrenched.
In a Darwinian twist, the survival of the fittest means that only a handful of system setups make bitcoin mining actually worth it. The economics of scale of bitcoin mining is conspiring against the network’s biodiversity. And this is happening as we speak. A recent press release from ghash.io shows that around 45% of hashing power comes from BitFury ASIC based miners. 
A sophisticated attacker would only have to subvert a handful of different standardized setups to compromise the network.
If a single entity were to wage a zero-day attack and compromise just a handful of ASIC based miners, they would create the biggest bitcoin botnet ever known to exist.
Is the Bitcoin ecosystem in danger of a 51% attack? It very well may be—today.