Application of mining hardware to future blockchains.

[Witrebel]

It seems to me that a good deal of press and research is covering blockchain technology and pointing to future applications which would presumably deviate from the Bitcoin protocol/blockchain. 

As I understand it Bitcoin derives its security from the millions of dollars in hardware currently deployed and hashing around the world.  Correct me if I am wrong, but this hardware simply hashes "work" aka data that it is fed from a network, and attempts to find a nonce that solves the block against a target provided by the network.  This is why you can point a (SHA-256) miner at any compatible (SHA-256) alt-coin, and hash on that blockchain.

There seems to be a great deal of debate regarding the future of Bitcoin.  How to handle microtransactions, or is the protocol even meant to do so.  How to scale the protocol regarding block size.  It is conceivable that a third party, be it private, public, corporate, open source, whatever... might roll out an improved protocol separate from Bitcoin, and it is possible that this new protocol may overtake Bitcoin.  Whether you believe it is probable or not, for the sake of discussion, lets pretend that happens. 

My question to the real guru's out there is two fold:

Is there any thing specific about the way ASIC miners handle work that is tightly tied to the Bitcoin protocol, as opposed to some other protocol implantation that uses SHA-256 and nonce based POW?  Aka, could some vastly different protocol with entirely different source code from Bitcoin still benefit from the established hashing power worldwide.

Second question is:

Does the Bitcoin protocol REALLY lack that much at a fundamental level, or could it be added to and forked to retain supremacy?  Is there room for another protocol to do things differently and end up with a more adopted/useful blockchain?

Third question:

Do you think this is a likely scenario?  And if so, do you think the new protocol would be intentionally designed to be backwards compatible with existing mining gear, or would it be intentionally designed such that it required new mining hardware.

[saturn643]

Is there any thing specific about the way ASIC miners handle work that is tightly tied to the Bitcoin protocol, as opposed to some other protocol implantation that uses SHA-256 and nonce based POW?  Aka, could some vastly different protocol with entirely different source code from Bitcoin still benefit from the established hashing power worldwide.

There is nothing specific about ASIC hardware that ties it to the protocol. However there is something specific that ties it to only be able to perform the SHA256d (SHA256 double) hash. The hardware is literally hard wired to only be able to perform a SHA256d hash function on whatever data it is fed.

Does the Bitcoin protocol REALLY lack that much at a fundamental level, or could it be added to and forked to retain supremacy?  Is there room for another protocol to do things differently and end up with a more adopted/useful blockchain?

First of all, the protocol is different from the consensus rules. The protocol is simply just the formatting of data and how it is transmitted. What people actually care about are the consensus rules, and those are what define what is valid and not. The consensus rules can be changed, protocol rules not really. Changing the consensus rules will typically involve a fork.

Do you think this is a likely scenario?  And if so, do you think the new protocol would be intentionally designed to be backwards compatible with existing mining gear, or would it be intentionally designed such that it required new mining hardware.

There are guaranteed to be changes to the consensus rule. Most of the BIPs involve some sort of change to those rules. The only thing that would require a change of mining gear is to change the algorithm used for mining. This would only happen if SHA256d was found to be unsafe and broken.

[Witrebel]

So as I thought, the existing mining hardware would be very happy mining for any protocol that requires SHA-256d hashing.  This could be an entirely new protocol, completely separate from Bitcoin.

I think it is quite likely that there will be a Bitcoin "challanger" that arises from all of this block chain attention. 

My question still stands, are there any/many inherent flaws in the Bitcoin protocol that would provide a challenger protocol to be competitive?

And if so, do you think that any new protocols would attempt to retain backwards compatibility with the existing SHA-256d hashing hardware?

[saturn643]

So as I thought, the existing mining hardware would be very happy mining for any protocol that requires SHA-256d hashing.  This could be an entirely new protocol, completely separate from Bitcoin.

I think it is quite likely that there will be a Bitcoin "challanger" that arises from all of this block chain attention. 

My question still stands, are there any/many inherent flaws in the Bitcoin protocol that would provide a challenger protocol to be competitive?

And if so, do you think that any new protocols would attempt to retain backwards compatibility with the existing SHA-256d hashing hardware?

It all depends on what you define as a flaw. There are some features that Bitcoin doesn't have that altcoins do. There are some competitive altcoins out there, but Bitcoin still holds the highest market cap.

I highly doubt that they would continue to use SHA256d since the existing hardware can be bad during the beginning of that coin. Current ASIC hardware can mine a coin in its early days and push the difficulty incredibly high. That can immediately kill a coin if no one can mine on it. Also, that would lead to one of the potential flaws, mining centralization. SHA256d is easy to put into an ASIC so it becomes much more expensive to run mining rigs. A lot of people would prefer that everyone can mine using CPU or GPU and not ASICS since CPUs and GPUs are easier to get and much much cheaper.

[Witrebel]

It all depends on what you define as a flaw. There are some features that Bitcoin doesn't have that altcoins do. There are some competitive altcoins out there, but Bitcoin still holds the highest market cap.

I highly doubt that they would continue to use SHA256d since the existing hardware can be bad during the beginning of that coin. Current ASIC hardware can mine a coin in its early days and push the difficulty incredibly high. That can immediately kill a coin if no one can mine on it. Also, that would lead to one of the potential flaws, mining centralization. SHA256d is easy to put into an ASIC so it becomes much more expensive to run mining rigs. A lot of people would prefer that everyone can mine using CPU or GPU and not ASICS since CPUs and GPUs are easier to get and much much cheaper.

Right, as I recall that was the intent of Scrypt, to be more ASIC resistant.  And I would agree that anything coming directly from the open source community would probably attempt to retain and emphasize any aspects of a protocol that would lead to more decentralization and keep mining accessible to the vast majority of the public. 

The scenario I am envisioning is more along the lines of central banks and technology leaders with established credibility trying to offer a more main stream block chain / protocol.  As I see it bitcoin is currently lacking widespread mainstream adoption.  Furthermore it seems that perhaps Bitcoin may have issues scaling to the usage levels that micropayments and high volume small dollar transactions would create.  E.g. if it became the next visa/mastercard/paypal it might not perform as expected due to either fee's increasing or clogging of the mempool. 

So basically, im proposing that someday we might find ourselves competing against some other protocol for supremacy as a crypto currency.  If that "alt-coin" has mainstream backing, it just might have a chance.  Ergo, If bitcoin becomes number 2 will existing Bitcoin miners be hosed? Or would it be prudent to base any new protocols on the existing ASIC hardware? 

Or for example, if Bitcoin crashed and burned, and a few years later something else rose up, would that "something else" be likely to use SHA-256d?

[kano]

...
Right, as I recall that was the intent of Scrypt, to be more ASIC resistant.  And I would agree that anything coming directly from the open source community would probably attempt to retain and emphasize any aspects of a protocol that would lead to more decentralization and keep mining accessible to the vast majority of the public. 

False.

Scrypt was a scam to convince people to CPU mine while 'they' GPU mined and made lots of scrypt coins
(I even argued back then that the scrypt setting were too low to stop GPU mining ... which was obviously correct)

SHA-256d?

It's a special variant, in two ways, of a double SHA-256 since it doesn't need to know the full answer and thus is optimised a few % faster.
For most ASIC, the input is also after the first SHA-256 has been done since that's the same for each 4-billion nonce checks.

Go look at any GPU implementation and you'll see.

[Witrebel]

...
Right, as I recall that was the intent of Scrypt, to be more ASIC resistant.  And I would agree that anything coming directly from the open source community would probably attempt to retain and emphasize any aspects of a protocol that would lead to more decentralization and keep mining accessible to the vast majority of the public. 

False.

Scrypt was a scam to convince people to CPU mine while 'they' GPU mined and made lots of scrypt coins
(I even argued back then that the scrypt setting were too low to stop GPU mining ... which was obviously correct)

SHA-256d?

It's a special variant, in two ways, of a double SHA-256 since it doesn't need to know the full answer and thus is optimised a few % faster.
For most ASIC, the input is also after the first SHA-256 has been done since that's the same for each 4-billion nonce checks.

Go look at any GPU implementation and you'll see.

So you're saying scrypt is a variant of SHA-256d, due to the aforementioned features, but clearly it is not resistant enough to optimization to keep GPU and ASIC mining from overtaking CPU mining?

Again, my root question is... If the mainstream tech sector attempted to take on Bitcoin with an alternative protocol, would it be likely to retain compatibility with SHA-256d miners? 

[kano]

No, that's 2 separate answers I gave.


1)
Scrypt is scrypt, not sha256.
However it was initially designed so that the designers could scam people by getting everyone to CPU mine, while they GPU mined much faster.


2)
The bitcoin sha256 is not a standard full double sha256.