Topic: 51% attack threat (Read 1089 times)

Thanks

Potential?

Firstly, consider the costs:

Cost for Cluster GPU Instances runs $2.10 per hour, or $1.70 after maximum discount for massive purchase volumes.
 - http://aws.amazon.com/ec2/#pricing

Revenue from Bitcoins produced by Cluster Computer Instances runs from $0.02 per hour on down.
 - https://bitcointalksearch.org/topic/m.716370

So economically, that's suicide.

But lets say, for whatever reason to prove this you could spend a half million dollars for the day and Amazon somehow happened to have enough excess capacity (which is unlikely they would).

So what does the ability to do a 51% attack give you?  

Sure, you can refuse to include all transactions and grief bitcoin for a while.  You probably aren't going to have enough excess hashing to be able to go back and rewrite the blockchain even a block much less two.  So anyone with several confirmations is not at risk of any double spend.  And because the network hashing level went up so quickly, anybody doing a transaction is going to be extra cautious ... waiting at least 6 confirmations, maybe more.

Eventually (after a few your hours) your budget runs out, and everyone continued to mine and now extends off of your last block, and life continues.  Included in the next several blocks will be all those transactions you refused to include.

So ya, ... technically bitcoin is at risk of being griefed by a party with massive hashing capacity.     But if it comes, it isn't going to be from Amazon's GPU clusters.

Amazon cloud isn't efficient enough at mining to do a 51% attack.

CPU's are useless, thats 99% of their computing power that can't really be used for mining. They have a smaller GPU cloud, but again, its not efficient enough or large enough to do a 51% attack.

No.

Doesn't the amazon cloud service pose a potential 51% attack threat?