Quote from: dagelf on November 13, 2014, 07:10:11 AM

Quote from: electerium on November 13, 2014, 05:04:02 AM

Moore's law is already cracking, from the perspective that IPC increases out of AMD and INTEL have been strictly nominal for the past few years. So from that narrative, which I think also holds strong (looking at the supercomputer list when you normalize some combination of flops/kw or cores/kw) in the proprietary chip market also (e.g. IBM).

So what's going to likely happen moving forward is that sha256 gets slowly eroded, which is basically what has happened to every other industry standard cryptographic algorithm.

In the ASIC chip industry, people are already down to 28 and 20nm. Soon enough (e/g 1 year) when everyone in the industry has reached down to 20nm you'll see a plateau in computing power between chips. The competitive advantage will dissipate between manufacturers as everyone optimizes their chips at 20nm.

So the point is that there is basically no likelihood of a zeroday event where someoen ramps up enough computing power to brute force out sha256 tomorrow.

Again, isn't this just addressing the "front-door" approach, that everyone seems to stare themselves blind at?

Let's try this differently, how centralized are these pools?

Discus Fish
GHash.IO
KnCMiner
AntPool
(https://blockchain.info/pools)

What will it take to take them out, and if done, how long will they be down for?

Another scenario - how elaborate a hack will it take to link them together to do a 51% attack to empty some big wallets?

How many layers of security would you need to get through? How many stolen ssh keys will it take?

Quote from: dagelf on November 13, 2014, 07:18:43 AM

Quote from: franky1 on November 12, 2014, 01:24:46 PM

Quote from: dagelf on November 12, 2014, 11:31:29 AM

5) Add to that security. Who holds this metric: How secure is each GH?

restored to a previous point.. ?? im not even going to comment on the stupidity of having restore points as thats the fungability argument. bitcoin will continue on as it should, as you say it would require a hell of alot of things in combination to cause the bitcoin ledger to be re-written and governments would not waste resources or risk fiat catastrophe based on bitcoin. after all there are over 100 FIAT currencies.. do you see america trying EMP explosions in russia and china to stop the BRICS development.
as for your comments on how secure is GH.. well better than KH better than MH, but not as good as TH and definetly not as good as PH.. so dont worry about the small stuff we are over 25% nearer to EH than dropping all of the way back to GH

What I mean, is, how easy is it to gain access to the largest mining operations and pool control structures?
On the physical layer?
On the OS layer?
On the social engineering layer?

What I mean is, how easy would it be to disrupt the 5 or 6 largest pools, and take half the network hashrate offline? How feasible is it? And as with anything, surely a cost can be attached to that... and a reward. Does it even out? Has anyone done a qualitative calculation? At what price point will it become feasible?

Where is this calculation? Or do I have to do it myself...?

For someone claiming to have read into the matter and know a lot about network security you certainly don't seem to know too much about the Bitcoin protocol. As I've answered before, taking down those pools would technically be the best possible thing you could do for the health of the network, since within hours all miners will step over to DECENTRALIZED solutions like p2poool, since none of them want to be missing out on potential mining profits. I wrote this before but I doubt you read it.

Now lets step over to the social engineering part of it all. Ok, so congratulations you've hacked into all of those pools and you can now unleash your evil plans on the blockchain... One problem, you can't STEAL anyone's wallet! WALLETS ARE SECURED BY CRYPTOGRAPHY STRONG ENOUGH THAT IT WILL TAKE MORE THAN THE ENERGY OF THE ENTIRE SUN FOR ITS ENTIRE LIFESPAN TO CRACK. Having 51 or even 99% of the network changes nothing about that.

You could prevent transactions of your choosing from gaining any confirmations, thus making them invalid, potentially preventing people from sending Bitcoins between addresses. You could also reverse transactions you send during the time they are in control (allowing double spend transactions), and they could potentially prevent other miners from finding any blocks for a short period of time. Thats really about it

And all of this would be blatantly obvious to people monitoring the blockchain. Miners that notice they're mining on a malicious pool would step over to a different pool, or simply to p2pool. Once miners step over you and your evil plans will be left in control of nice pools.... controlling exactly 0% of the networks hashrate.

There is nearly no profit is obtaining 51% of the network through hacking except for doing some doublespends after which miners stop providing their hashing power to mine for your evil plans. In short, you'd be doing the network a favor since people will finally step over to decentralized mining pools.