...
The intent of LTC when released was to be resistant to GPUs/FPGAs/ASICs for the foreseeable future. One quickly went down, and with the advance in lithographic technologies and geometry creation, the second is being bypassed and it's going straight to the third one.
As a note, the thread on the LTC forum is not from the "developers" who want to fork LTC. So it still looks like profit motive. Maybe this is the new paradigm that people are going to use to make money now: forking already existing coins to new algorithms with the intent of "ASIC-resistance."
I will say, ASICs are just a new starting barrier for a coin. In no way does it decentralize the system as anyone is still able to participate in the system. In fact, ASICs provide the service of securing a blockchain which is just as important. After all, the same arguments happened with CPU vs. GPU. And in fact, for a long time, GPU was akin to ASICs as only certain cards could mine more effectively and we had a whole bunch of hobbyists trying to outdo each other by tweaking existing components to gain an additional edge. ASICs to me are just the next evolution of that. The difference is that instead of manipulation from the hardware level, it's being done at the wafer level.
You should drop this post in that thread. I don't think I agree that ASIC secures a blockchain. Once a quantum computer comes out any blockchain is in danger and centralizing with ASIC is the wrong direction in my mind. Think On this, any large enough government can and may use ASIC tech to centralize hash and take control (Bitcoin better watch for this). It can be done now with the Us gov if the will was there.
I may drop that post in that thread if it gains further traction.
On the realm of quantum computing, that is a fair concern. After all, quantum computing that resolves both entanglement and the eigenvector trap (both a physical and a code-based restrictions) would easily simplify all of the integer-based algorithms and make solving them relatively trivial. Instead of taking 2^O(n) time, that's reduced to 2^O(log n) time (from exponential to polynomial) in the best case scenario [and at worst, with a bad implementation, we would have something in quasi-polynomial time].
Once that happens, any crypto-currency that is secured with an algorithm that runs in exponential time will be well...fucked. So even a switch to X11 and so forth would be moot. At the minimum, in order to resolve this, the new algorithms will need to have 2^(2^O(log n)) time, i.e. double exponential run time so that at the best, Shor's Algorithm will reduce it to something that is still NP-time. [It would be fun to see a blockchain secured with something like Akerman's function though...]
In this sense, ASICs are a better security mechanism than what we have right now. I still disagree that ASICs will have the centralizing affect
if enough people participate. I'm still of the opinion that if the government wanted to take down a coin like Bitcoin, they could easily do so by implementing their own sort of ASIC chip, and having a fab or foundry produce/assemble them. Of course, now it's actually significantly more expensive then it was prior to the advent of ASICs for the SHA-256 algorithm [based on current network power, lithographic advances in the realm of ASICs and continuously improved adaptation, it'll likely be a ~$500 million proposition right now, back in mid-2012 while ASICs were in the development stages, it would have only been a ~$35 million proposition].
