We were able to come up with a basic architecture for Equihash ASICs that would be able to successfully follow a hardfork that chose any set of parameters.
1 Like
When i wrote the last weeks again and again that changes and tweaking mostly wonāt be enough i was offended as a bitmain agent, bitmain employee, bitmain promotor, dumb, whatever not.
At least nice to see that my arguments and thoughts that the tweaking/changes might not be enough are backed by an unbiased personen now.
https://news.bitcoin.com/siacoin-developer-asics-are-money-printing-machines-for-manufacturers/
Insights from the Siacoin developer, a lot has allready been said in several topics regarding the Asics 
The new medium article by David Vorick has completely shifted my views on ASICs. He makes a solid case that ASICs are much more flexible than many GPU mining communities believe and can be made adaptable to hard forks. He also makes the case that Bitmain is in a solid position at the top of the ASIC manufacturing market and has or could very likely throw their weight around to stay there.
My initial thought was that we canāt move to PoS fast enough, but my next thought was that Iām on the fence about PoS in the first place.
If the hashing algorithm were to change I think Ethash would be the best move for the next few years until PoS was more understood, stable and widely used. This would allow GPUs to continue mining alongside existing ASICs.
My thoughts continue to evolve on this topic.
1 Like
this already is happening, SIAcoin https://sia.tech/ is developing their own ASIC for decred algo, they were beat to market by bitmain, but they blog and release in depth articles on their journey/discovery of the ASIC ecosystem and spectrum.
is a very detailed blog they release yesterday about the inevitability of ASIC mining, and Bitmains near insurmountable dominance of the manufacture. They also detail the ease at which bitmain and other asic manufacturers can make asic that are highly adaptable to any minor equation changes (such as N/K changes in equihash)
I just bought FPGA rig, not as efficient as ASIC but adaptable to any algo, and miles ahead of GPU in terms of hashrates and efficiency, before that I had ordered Z9, but now I am not so worried about fork anymore, just relaying info
1 Like
I am not on the fence about POS, it is complete centralization into the hands of the wealthiest. rumored stake for ethereum is 1000 eth, so unless gpu miners have at least ~750k USD laying around to put up as stake, they are ass out.
1 Like
Would the solution be to just lower the staking value for zec to something more attainable for smaller users? I think this number is hard to hit because of the exchange rate fluctuation over time. When it comes to mining, wealthier miners have an advantage due to economies of scale. PoS would remove those advantages.
The post was a good read but its worth bearing in mind:
- David is an ASIC manufacturer so his views on GPUs and ASIC resistance are not necessarily unbiased.
- Obelisk and Bitmain are competing against each other in the blake2b (Sia) and blake256 (Decred) markets.
- The claim āBitmain plays dirtyā was made several times but no evidence was provided.
1 Like
Your points are completely true - there was no evidence of Bitmain being at any fault for their chip not being manufactured.
The main shift was that I hadnāt realized how flexible ASICs could be: that some could be able to survive hard forks with minimal performance loss by becoming slightly more generalized.
Where did you buy it?
Davidās logic is sound if you have a money making machine you do not sell it until the margin of doing so is better than the margin of operating it. Example it looks like Bitmain just released an Antminer S9i for $1010 which is more efficient than the S9. A reasonable assumption that this represents the top of the S9 line equipment that Bitmain ran for its self and is now releasing the market as the S11 takes over their internal mining footprint.
The chips/rigs they have in R&D the would mine before selling. Very safe to say they paid back their full investment on the Z9 before opening it for sale.
Bitmain has choke points in the supply chain. Samsung and Apple do similar ātricksā. Manufacturing is a scale business, meaning it centralizes, whoever has the most capital/demand at the different points in the supply chain will dominate that ānodeā in the supply chain.
3 Likes
We appreciate the opportunity to be a part of the community and we are learning in the process. Thanks for the encouragement. I just put my name in for Zcon0, hope we can have more of this conversation in the meatspace.
1 Like
Refreshing to hear someone in the industry confirm what so many folk have been saying for months on the ASIC mining thread.
That consumers are unwittingly buying second-hand equipment passed off as ānewā; and that the mining industry is a natural monopoly with high barriers to entry.
Btw, Bitmain have responded to Davidās article which they are calling a āconspiracy theoryā. On Sia's āThe State of Cryptocurrency Miningā - blog.bitmain.com
1 Like
While they are mining i seriously doubt itās the hardware they sell later for several reasons:
-
First, iāam checking all my miners when they arrive, about 70% of them are 1st batch ones. I open everyone of them indeed, exactly due this rumour, and to some extend out of curousity of course. I do not care about the seal as it wouldnāt be anyway mostly useless to send them back in case a defect occurs having in mind the hazzles and cost for shipping forth and back, again customs, whatever not. Of course i might be lucky that in the last 30 years i never had a vaulty PC component or Asic, GPU, whatever computer related.
This said, i never have seen any traces that the miners are used. The latest X3ās and B3 are no expection there, just as new as new should be.
P.S.: I clean mine every 2 weeks and i never was able to clean them again such clean as they arrived and i put a lot of time into that not sparing them to have them offline and using whatever not, but after 2 weeks there are just traces inside the Asic i can not remove anymore. -
Cleaning a Miner to such step that it would look like new, i doubt, if possible at all it would take mostly a lof working hours and additional personal for all the steps needed to disassemble, clean, check, assemble the unit.
-
It would be as well some internal logistical problem, or maybe better, additional work. Removing a unit, replacing it with another one, setting up the ip again, and whatever not. While this is not a real argument and someone that wants to make it will make it. But having in mind how big Bitmain is i have my doubts that they set the priority for such action above it cons, like additional menpower, payments, wages, disruption of the working process and whatever not. I very good could imagine it in a smaller company, but at such levels like Bitmain ā¦ i just have my doubts.
-
The logistical point of view. When Bitmain ships they ship fast and a lot at once. There would be only one option to do so, weeks in advance taking the units offline to prepare and begin the refurbish. This time, combined with the additionaly costs for menpower and additionally costs would in my opinion not repay the lost mined profit for that time. Iāam not saying again itās not possible, i just have my doubts that from a such profit oriented company like bitmain they wonāt profit some extra dollars for the extra work needed for it. I bet Bitmain is better profit organized than this.
-
It would as well increase the chance of garantee claims in my opinion as used hardware for sure would have a higher chance of getting vaulty. Again nothing a manfucator wants to have.
-
It doesnāt fit with the argument that bitmain anyway uses better equipement, something i would be more willing to believe from a logical point of view.
-
While here and there are some claims that the 2nd hand sale happens, like the Sia/Obelisk guy, have in mind a competitor to Bitmain, i have not seen any real unbiased evidience or proof for it.
If i would have any such 2nd hand hardware shipped by bitmain i would share this fact, but as described allready, i did not happen so far and while most miners give something about the seal and their garantee, i donāt. Neither do i plan to resell them, nor to send back in case of a failure.
And again, iāam not saying it doesnāt happen or does not happen, but i think itās highly unlikely.
1 Like
"At launch, Bitmain introduced a strict one-miner-per-user batch allocation policy and other methods [1][2] in order to restrict hoarding and combat centralization. The number of miners made available per round was purposely limited to help further regulate their introduction into their respective ecosystems. Since then, weāve enforced this policy with the launch of all ASIC miners for cryptocurrencies that can be profitably mined by GPUs. The articleās calculations here are merely speculative. "
Clearly not true since the Z9 has a limit of 50 within 2 weeks after it was available.
1 Like
Today i got a Bitmain B3 Miner and itās a really interesting one inside.
Writing this because i think, or better it could be, the Z9 is on the same chip or architecture eventually.
As always i open the miner to see if they are used or not, due the widespreaded myth of Bitmain shipping 2nd hand miners. Itās a brand new one, not a piece of dust in any corner.
Itās an interesting design in my opinion and uses 4x Bitmain AI Sophon Neural Chips BM1680.
Itās the same chip that is in their graphic cards. Iāam not the super tech guy, but from what i see/read the chip supports up to 16GB DDR4-2666 RAM.
Is it possible that such modified design is used for the Z9? I think it would make sense with the Chips that are used in their Graphic Cards as well, not?
3 Likes
ePIC is agnostic (open) to whatever hardware format the community rallies behind. The purpose of our Grant application is to provide leadership and expertise on how the Zcash community can form its hardware roadmap. We will research ideas, starting with Equihash as that is the current standard, but as the community provides input from the software & governance perspective we provide feedback on how that translates into hardware.
1 Like
Keep in mind that is one-miner-per-user sale to the public at retail prices. Show-up with a few million dollars and I suspect the policy would change pretty quickly. Not to mention whatever went on before the public sale started.
2 Likes
This is a bit of a tangent, but Iād like to point out the risk of āreverse causation errorā in addition to the risk of ad hominem in comments like this.
First, the problem with ad hominem, is of course, that if a person says something, and there are reasons why they might be biased in favor of that thing, that this doesnāt mean the thing isnāt true! Thatās the fallacy of ad hominem. In addition to being a fallacy, it often leads to conversations going off the tracks of figuring out whether the thing is true (which is feasible and helpful) into figuring out whether a person is good/bad/loyal/disloyal/biased/unbiased (which is impossible, and trying is unhelpful).
Everyone is biased about everything that they say out loud. If they didnāt have any care, any knowledge, or any personal stake in that thing, then they wouldnāt be talking about it. Letās focus on discerning the truth and not speculating about other peopleās internal lives.
Now, about the reverse-causation bit: If you see that someone is engaged in activity X and they also have belief Y, then you might suspect that the former is biasing or influencing the latter. In this case Activity X is āDavid Vorick runs an ASIC companyā and Belief Y is āDavid Vorick believes ASICs are inevitableā.
But have you considered that it is actually Belief Y causing Activity X instead of the other way around? David Vorick wrote and spoke about the inevitability of ASICs for years before he founded his ASIC mining company. It seems obvious to me that Belief Y is the original cause of Activity X.
I see this kind of thing all the time. I personally, and the Zcash community as a whole, are often targeted by it. People often off-handedly say things like "Oh, of course Zooko thinkās Moneroās RingCT are insufficiently safe (Belief Y) because he is a Zcash developer (Activity X).ā. But in fact, I had Belief Y (for rings/decoys/mixins in general ā this was before RingCT itself) before I decided to devote a substantial fraction of the rest of my life to Zcash! So in fact Belief Y is the original cause of Activity X, not the other way around.
But, yeah, probably the best approach is just donāt try to include speculations about other peopleās motivations and intentions into the conversation. Itās not gonna work out well. 
2 Likes
Fantastic! Looking forward to seeing you there.