Bitcoin Forum

Just a thought. In technical terms, what would happen if someone with a bunch of ASIC miners pushed their machines onto a Scrypt chain? Would absolutley nothing happen to the netowrk? Would it cause unstableness in the chain? What would happen. Im betting that something has to happen. But thats just a thought..

Rumor has it that this has already happened (Scrypt ASIC), then they stopped mining, which is the cause for the projected 15% drop in difficulty next round.

Obviously, an ASIC for SHA256 mining cannot do Scrypt mining.   

So you are asking what would happen if an ASIC that does Scrypt were invented?

It probably depends on how the ASICs are put into play.  If they are intended to be units for sale to the public, then it might be something that would benefit the leading Scrypt-based alt.  If the units are intended to be used in a 51% attack, then it is lights out for whatever Scyrpt-based alt(S) they target.

from what i understand an asic designed for scrypt would be much less relatively effective as bitcoin asic's are

But the recent bitcoin history is teaching us that there isn't anyone out there that is both idiot and eccentric enough to try a 51% attack with asics...

Noo, just out of curiosity, what would happen if a person with say a 500GHS SHA asic pointed it towards a scrypt curreency. Probably nothing. But if so, what does it do? Cause stales for others, or anything at all.

Definitely nothing because a SHA asic is not compatible with SCRYPT.

I am actively looking into this. The problem is that ASIC's are not economically worthwhile for LTC yet. Although we technically have a high valuation, the markets are shallow- way too much slippage on the exchanges at the moment. If you guess that on the lower bound an ASIC might cost 100K usd to build (That's really low end too, one quote I have just for preliminary design is around 40K USD, just to design, and that's no guarantee of how well it hashes) and you mine 1 million coins super fast, you still have to sit around and hope LTC takes off to recoup your investment. If you sell those 1 millions coins, you would destroy the LTC market.

Chicken and egg really. But people are looking into it. A more reasonable start would be an FPGA but so far the results are not very encouraging.

I agree, if they keep the fact that they have a ASIC that does Scrypt secret, use enough to corner the market and perhaps a 51% attach, it will be lights out for that coin. Maybe the Point of Stake Scrypt Coins will be more resistant to the attack and survive. Who knows, but I am concerned about the future of Sha-256 and ASICs now. I'm glad that as far as we know Scrypt is safe.

It would make much more economic sense though, to keep the SHA ASICs a secret, and mine, rather than 51% attack. By keeping it a secret, they could mine ludacris amounts of valuable Scrypt coins, and make a killing on the market. However, it would be difficult to gather the funding to produce Scrypt ASICs unnoticed.

As I understand it, ASIC bitcoin miners won't work on scrypt because they do not provide a 'proof of work' for the solved block.  What if the ASIC found the solution, and then passed that solution on to the video card so that the video card can generate that necessary proof of work?  In other words, couldn't they be made to work in tandem?

An ASIC does ONLY what the electronic paths in the silicon allow it to do.  SHA256 hashing is what the ASIC in an ASIC miner does.   If even one tiny little change to the computation is needed, that ASIC design is useless.   Scrypt is way different from SHA256.  There is no computation from any ASIC miner hardware that is even remotely useful in performing a Scrypt hash.

Same thing that would happen if a person without a 500GHS SHA ASIC pointed it towards an scrypt currency. The only thing a 500GHS SHA ASIC does is SHA and whether you do SHA or don't do SHA makes no difference with an scrypt currency. SHA is not scrypt, so it works the same on an scrypt currency as anything else that's not scrypt, such as nothing at all, or addition, or whatever.

No. There is no distinction between "that solution" and "that necessary proof of work". They are the same thing. The proof of work is that you found the solution, proving that you did the work needed to find it. The whole scheme works because, and only because, the proof of work is inseparable from the block and transactions that are solved.

I understand that the video card has to go through the process to find the work itself.  But couldn't that process happen with a little nudge in the right direction by the ASIC -- in theory?  The ASIC is 'dumb' in the sense that it's getting the answer through hardware bit flips at extremely fast speeds.  It gets the answer, but has no work to show for how it got it.  I'm merely suggesting that it might be able to pass that result to the video card, and the video card could make some use of it.

You just made TRC cry by forgetting about its last month

If this happens, it would certainly freak out rig owners and investors, as to what happens to the chain, it would be disrupted and if the configuration is not retargeting at every block, it would make the difficulty be OLYMPUS with no water or ground to land in.

That is nonsense. 

You seem to not get this concept.

The "AS" in ASIC is APPLICATION SPECIFIC.

The specific application is a double SHA-256 hash of a block header and incrementing nonce.  An ASIC can't do anything else.  It can't point GPU in the right direction, it can't do part of Scypt.  Hell it can't even hash SHA-256 in a different manner for example a triple SHA-256 hash or a double SHA-256 hash of a different sized blocked header.  It can ONLY do one specific thing.   All the SHA-256 chips ever produced and ever will be produced combined have 0.0 hashes per second on a Scrypt chain.

No. There is no "nudge". You either have a solution or you don't. When you know you have a solution, you are done. If you don't know you have a solution, your work is of no use because it's millions of times more likely not to be a solution than to be a solution. The "work" is in determining if you have a solution or not.

The answer is the work to show for how you got it. The whole point of the design is that work is needed to produce the answer, thus the answer provides proof that work was done.

Until you have a useful result, you don't have a useful result. Once you have a useful result, you are done.

ASIC on scrypt would just destroy the purpose of its invention, to make it profitable for CPU users. So, if they produce in large quantity and starts selling them, scrypt would not be profitable for CPUs.

I think we can pretty much reason out now that that's impossible to sustain. If it's profitable to mine on CPUs, then more and more people with CPUs will mine. This will continue to occur, raising the difficulty, until it becomes unprofitable for most people with CPUs to mine. Then only those people whose circumstances make them especially efficient miners, such as a very low cost of electricity, will find mining profitable. This is an inescapable economic reality.

It would actually be very beneficial for Bitcoin if a scrypt ASIC was made and used to attack the alts, they could all be destroyed!

I think you might be exactly wrong on that.  Bitcoin benefits greatly from the alt market. 

However, if someone can design a efficient asic in the future, the difficulty will skyrocket like what happened to bitcoin. Highly impossible and i quite agree with your point but in the future, who knows? Maybe someone would come out with one.

you can't mine scrypt with a sha-256 asic  ::)