[cryptography] Bitcoin in endgame

coderman coderman at gmail.com
Thu Feb 23 05:25:47 EST 2012


On Thu, Feb 23, 2012 at 1:24 AM, Eugen Leitl <eugen at leitl.org> wrote:
> On Thu, Feb 23, 2012 at 01:00:20AM -0800, Jonathan Thornburg wrote:
>> On Thu, 23 Feb 2012, James A. Donald wrote:
>> [[for attacking bitcoin]]
>> > botnets cannot compete with legit miners, because to get a reasonable return,
>> > you need to mine with a graphics card, and while mining with graphics card,
>> > your graphics goes to hell, which will cause the most slow witted owner of a
>> > zombie computer to do something.
>>
>> It's pretty easy for the botnet to monitor for human activity (e.g.,
>> mouse movement, non-screensaver processes running) and suspend their
>> activity then. ...
>
> What is the actual problem with botnet blocks (which, so far
> I'm aware of, is a purely theoretical situation at this point)?

not theoretical; roughly 10% of the current capacity is allocated to
"mystery miners". the pool operators know that bot miners exist, some
take steps to thwart their participation in their pools. others don't
mind so much.

OpenCL/CUDA kernel tuning allows a workload to run in parallel with
standard desktop operation sufficient to be negligible. flash content
takes more resources and imposes a heavier footprint than a stealthy
miner. all this before you even get real sneaky, like reacting to user
activity.

what has been most fascinating to me over time is just how malicious
and skilled various actors have become within this ecosystem. every
peer their own bank, yet hardly protected like one...



More information about the cryptography mailing list