If there was a drop like that it would be noticeable to anyone who is mining since their expected yield would go up considerably.
Coinwarz has the difficulty at 7303654.73
If there are miners using ASIC hardware on ZCash, it would be safe to bet that it’s very limited. ZEC difficulty has grown at a stead rate; c.f. Dash where it’s easy to see when ASICs took over.
NB: The mild inflection point on the ZEC graph between April and July is due to the JP Morgan news on 22 May and the onrush of miners jumping in following that positive news…seems soooo long ago, too o.O
@zooko - that’s what will happen to ZCash if you don’t stop it.
Ok, its not an easy problem to observe because of the different viewpoints, so lets recap the current situation
There is not enough evidence to show that asics are mining zcash yet
The odds of a Zcash asic currently being developed is high
There is evidence to show that asic miners have come into play with certain other cryptocurrencies recently (physical or otherwise) which further supports the previous statement
From these 3 observations we can conclude the only difference between a gpu mined Zcash (now) and a gpu/asic mined Zcash (soon) and an asic only mined Zcash (later) is time (1,2,3)
And thats how its gonna go, every road leads to this eventuality, its just the timeframes in between we can change, here the options (sorry if i missed any, which im sure of)
Asic resistance fork
Pre asic chip R and D / mfg distribution
Do nothing
Mining algorithm changes
Asic resistance fork
Sure we could fork, but then the fork only repeats the cycle, original zcash was GPU, then both, then Asic (1,2,3) (like monero and others)
And then every fork of every coin worth anything (1,2,3) (asics coming soon)
It happened with Bitcoin and forks of Bitcoin and now forks of those forks
Forking for the purpose of resistance only makes it sorta go away for awhile in a sense that its not you’re problem anymore, for now (out of sight, out of mind)
You’ve solved nothing and really only shows a lack of applied thinking from emotionally driven decision making
It will schism and weaken the network
To willfully stymie what may only actually appear to be an adversary because of fear that it may affect your profits is not the reasoning of a scientist but of a miner and is really only in a miners interests (no one elses profits are really threatened)
Secondly, its unethical (if you dont think so, I’ve got 27 synonyms)
Purposely undermining another business practice, regardless of how we feel about them, be it accurate or otherwise, is certainly not befitting the mission
We are a scientifically driven community and as such we must abide by ethical standards
We may be “keepers of the code”, but we cant say who cant or how, its open for all to use and have and mine however they discover to
(Those who would deny freedom to others deserve it not for themselves-Lincoln)
Pre asic chip R and D/ mfg distribution
I think this is a good contingency idea to have the design at least ready and mfgs (who agree) ready to produce if and when its given to them, to try to spread the cheese more if you will, but this is only an after the asic fact fix (1,2), and
If it was given to them sooner (and no other actions taken) it would accelerate gpu mining infeasiblity alot, I mean all gpu Zcash miners call no joy (1,2,3))
This is a one way decision (option limiting) too, there would have to be a lot of money sunk into both development and then manufactured by multiple manufacturers brought on by a zcash community consensus which is why I feel the design should be maybe created but held out until all of those options are already gone anyways
Do nothing
This is the Bitcoin (and others) route, eventually asics will overtake gpus and then it will all be Asic (1,2,3), similar to above (again, timeframes)
A lot of the things discussed on this thread are about 51% attack, which is feasible although impractical
For example, lets say i have well over 50% hashrate on a coin from asics i developed and built (because I run a chip plant or something) which cost millions, and Im making millions and will be for a long time
The idea that I might do anything to jeopardize that is completely illogical (put lightly)
because you’re buying most of the coins from me anyways (or more accurately my employees who also depend on the network working properly lest they have a job)
because the longer it hashes smoothly and stronger, the more confident people are
because their money builds more chips, repeat
Although impractical, it is feasible so it shouldn’t be ignored
Mining algorithm changes
I believe this to be the most practical option for any sort proactiveness towards increasing gpu viability and 51% attack prevention and probably what we should discuss
Remember, its an unwinnable hashrate fight (in an actual hashrate sense) and so far weve discussed turn tail and run, and lay down
In a lose/lose situation, the best you can realistically hope for is a tie
The only option is adaptation through some mining algorithm manipulation.
There are other ideas we should explore as well
I like the daisy-chain idea (asic,gpu,asic,gpu every transaction or block which there are slight differences), i think it offers the most benefits, attack resistance,
continuously lower gpu difficulty vs other algo increasing hasrate (more and more asics),
higher gpu incentive (this would be due to the effective hash rates of both algorithms requiring being about the same to achieve 2.5 min block) difficulty growth self moderation,
network strength,
50/50 reward split across both algorithms (2 ways to mine it, seems fair, this where the contingency should/may be i.e. no closing Pandoras Box once its open)
But even this way will most likely eventually lead to a completely Asic driven zcash but I think this way buys more gpu viability time without (or less risk) potentially damaging the system, mission, community
Technical analysis of the actual feasibility of such a mining protocol would need conducted
I don’t know if that would be something to propose to the foundation or done by the zcash company (i assume the latter, maybe voting through the foundation to have it done I don’t know)
I agree - The daisy-chain , multi-hardware, multi-algorithm strategy sounds like the best option. A design with allows for the non-ASIC algorithm to be “plug and play” and easily changed out would be the best strategy, which would hindrance the possibility of an ASIC being developed for the GPU mineable blocks/chain portion of ZCash.
Comprimise, survival, i don’t think theres a perfect solution (C’est la vie)
Half the block rewards for increased gpu mining time at lower difficulties whilst the network total grows ever faster from asics vs. fewer and eventually no block rewards from gpu mining being squashed (and quickly) whilst the network grows ever faster from asics
Swapping one algorithm for another would require more time as people would be left out, just doesn’t make sense to axe so much computing power when a mutually (or even equally) beneficial coexistence could exist
Edit- assuming its done preemptively, gpus mine both so only compromising when asics arrive
PoW exists in these sort of protocol for a reason, We should value this the most.
And that’s about control on driving the engine of this protocol. It’s given to computation power.
In my opinion, Least important thing is complexity of computation. And main characteristic of the computation is that, No one should be able to guess an answer unless computing one. Most of hash functions have this feature as far as we know.
My thinking is what if there was a way to invent a PoW algorithm which is the most friendly to asic while having that characteristic. Meaning designing an asic for it would require the least complexity.
I’m not a cryptographer nor a mathematician or an electrical engineer. So I’m not qualified for this task.
Also it’s best to be unique for each coin. bitcoin cash is significantly less secure than bitcoin. making it more vulnerable for 51% attack. Though i don’t this is going to be a big issue unless it becomes.
If zcash foundation did open a project for inventing a new PoW and open-sourcing the manufacturing designs (letting everybody to use it). I will be happy
Also by keeping rewards for possible optimization can keep the open-source project competitive. @zooko How can we ask zcash foundation to fund this sort of thing?
Maybe the algorithm is blake2b since i heard it’s the fastest. Then by changing it a bit to not be compatible with siacoin it’s fine to use.
We need check if there’s consensus for this change.
Having GPU’s to mine an specific computation is probably waste of technology. These devices are intended to do general processing.
Most of us wish for crypto-currency to grow into our society. If it gets that big. No government can prevent mining. All they can do is to regulate it.
Are we going to have a poll for this or just keep discussing it? It needs to be decided fairly soon so ZcashCo can plan ahead if necessary. We should select from increase existing equihash parameters, myriad mining approach or do nothing. I am for increasing the memory usage and after that myriad mining if an ASIC is produced. If the memory footprint is increased to 2.5 GB it should buy a lot more time at the least and it will make mining more efficient.
We need to continue discussing for now, Zcash co is busy with sapling, but at some point i agree that a poll should potentially be conducted across the various channels once a more concise picture of the options and their ramifications is constructed
Edit- much to discuss, for example (and this only one idea) with the daisy chain concept, and the assumption that Asic miners mfg would increase over time, by limiting their potential profits (to an equality) may actually help decentralisation by forcing them to sell some asics (?)
If the security properties of that PoW are orthogonal to those of the ASIC-mineable PoW, then having both of them in required quantities gives you the benefits of both. As one example, a 51% adversary would need to be somewhere between 51% of both PoWs, and 100% of one and 2% of the other. That adversary needs to either be as diverse as the PoW set in their operations / expenditure, or so overpowered in one PoW that we’d have problems no matter what path we take.
Here’s some food for thought, mention zcash about halfway but mostly about vertcoin, it doesn’t seem to provide anything new in relation to whats being discussed here
I think other coins should just use Vertcoin’s strategy that the article mentions. Vertcoin always has algorithms “on standby,” ready to go in case of ASIC interference, and they can swap algorithms at any time without much difficulty. I would suggest talking to their team and get some input on strategy. Their team is available on Discord.
Monero recently did the same thing and we should probably go over it again but like I said, no new ideas, Verts already decided they’re not even gonna try to figure it out later by giving up every ship they think may be sinking and regardless of any new evidence which may present itself later to the contrary i.e. maybe it was a mistake but too late, its some kind of despotic law thing we whipped up in hurry to intentionally subvert, sorry
i like to think were maybe smarter than that
Edit- stand by algorithms are a good idea in case something gets broken, serves the adage better too many than not enough
When an algorithm is changed, what is the extent of it’s impact? I know it affects 1) Mining software, 2) Core POW code, but what else does it impact that needs to be rewritten/modified?
Another question is , does this always require a hard fork, or it can it be soft?
But to answer your question, I believe it would be a soft Fork ( where 2 or more different algorithmic types of mined zcash are compatible on the same chain) would be advantageous to a normal S F because neither side would ever be able to effectively out hash the other
Actual hashrate, oh yea huge difference but always 50/50 so to my understanding would eliminate a HF potential caused by one side effectively dominating the other
HF is zcash splits into 2 different zcash types now incompatible because of different chains, like bitcoin cash and btc
I think we need to educate a lot of people that you only need a soft fork to change a POW algorithm. - there is a lot of confusion and a general assumption by most that a POW modification would require a hard fork. Thanks for reinforcing what I thought to be true…
Any change to the current PoW requires a network upgrade (i.e. hard fork) because existing nodes will reject any block not using Equihash 200,9.
Perhaps a separate discussion thread can be created for those who want to research schemes such as Open-Source ASICs, Myriad Mining, Velvet Forks, etc.
Those topics, whilst interesting, should not block or distract the community from the key decisions under discussion which (at least originally) focused on Equihash and ASIC resistance.
Namely:
Status of the social contract between the community and the project to maintain ASIC resistance, which was the raison d’être for deploying equihash in the first place.
