Even with Memristor technology it would take more than a small group to jam out the global hashpower. There's billions in infrastructure and it would require trillions to topple.
As said before, the "small group" will be the top 4-6 miners, who already have enough power to overcome all the others.
That is a paradoxical thing about the bitcoin protocol: no matter how massive he mining network, there will always be a potential enemy with the all power needed to take control of it. Just as, no matter how big of an army a country has, it will never be big enough to protect it from a military coup...
Edit: Lets also not forget that 51% is where it becomes possible, its still a hard thing to do and to pull off succesfully would likely require much more than that.
Yes, they would need to have somewhat more than that (say, 65%) for extra safety against "heroic miners" bringing up their uneconomical miners to fight them. The cartel would also have to divide their resources between jamming the old chain and mining the new one, constantly adjusting the split so as to retain a majority in both.
However, once the "attack" is announced, many "weak soul" miners would surely switch, followed by the "semi-starving" miners who cannot afford the blockade; so the cartel's position will become more and more comfortable as the "attack" progresses. Indeed, if the cartel plays the PR well, there may not be any disruption: most everybody would upgrade before the deadline.
I thought the 51% attack was related to transactions. To change the halving you would have to change the bitcoin client. Those "bad miners" would be mining an alt.
Yes, it would be an altcoin that starts off with the state and history of the bitcoin blockchain, and uses the same keys so that all bitcoin users are automatically users of that chain also.
Anyone can create such an altcoin, by making a few cosmetic changes to the standard software so that the messages can be distinguished. After publising the 25 M software, a cartel with 51% of the hashpower is able to render the old 21M chain unusable, by orphaning all blocks found by other miners and ignoring all transactions directed at it. Therefore they can impose any change to the protocol (such as the 25 M limit) that would be less damaging to the clienst than the jamming itself.
If what you say is true then why has it not happened yet?
First, similar forks happened twice already. In both cases the developers got most of the miners to agree to the fork, even though it meant rewinding the blockchain by several dozen blocks. The excuses were quite good, so there was no dispute.
As said above, it is risky to undertake such a move without a comfortable margin, and without a good discourse that would lessen the PR damage. The cartel members must also stick together for the duration of the "attack". These conditions did not exist before. (It is only recently that the top 51% became all-Chinese.) Those condition may well exist one year from now. The motivation, for the top miners, is huge.
Poll
