Let’s talk about ASIC mining

And they wouldn’t unless they want to commit financial suicide.

That is true but that makes coins no different then VIsa or master card just with servers in China.
ETH is only real decentralised coin now even it was heavily premined.

Monero has potential if it can stay ASIC resistant.

1 Like

Good thing is that now there are three companies that can manufacture ASIC miners and use them for their profit, so Zcash server power can be seperated.

But I see your point.

True but the Z9 is less than half the cost now with the price drop. It is less efficient, but with cheap electricity the Z9 is the clear best choice.

Do ASIC manufacturers mine secretly? David Vorick of Siacoin (aka Taek42) implies this is happening:

For every miner you buy from Innosilicon, they can make two more for themselves and also pocket $8,000 (estimated). 60% of Decred for the first few months of ASICs was mined into a single address. Those ASICs we’re also from Innosilicon. https://www.reddit.com/r/zec/comments/8pbmsi/new_asic_from_innosilicon_hashes_50k_solss_at/e0ck0md/

This could mean that 100% of the possible PoW block reward was distributed to one single address, since Decred uses a hybrid PoW/PoS system with a block reward being split 60% (PoW), 30% (PoS), 10% (dev fund): https://docs.decred.org/mining/proof-of-work/. If he meant to say 60% of the 60%, that would be 40%, still a large amount to be in the hands of one entity.

4 Likes

Does anybody know if Nicehash can still be used to attack if Zcash updates the parameters to use 4.5GB of vram? Don’t the hash rental companies like Amazon have K80’s, which have enough vram, but what about nicehash?

Yes but im a small miner as i said above thats why .

…However, our network architecture shouldn’t be reliant on the trustworthiness of specific humans, but instead should be based on cryptography and economics…(c) zooko
Crypto was designed to work even if it would be banned by every government in the world.(If these governments don’t ban PCs and Internet) It’s really important quality especially now when Japan bans privacy-based crypto like Dash, Zcash and Monero. What will be with Zcash if Chinese government does the same. I don’t want to say, that they WILL do this, but they MAY do this. They can make Bitmain do things that may look like financial suicide but essential for company surviving under pressure of “national security” governmental institutions. They already control Dash, Monero escaped from them. What will the third privacy-based coin do? Surrender or no? Here is the main question of this discussion .

These Innosilicons are true “killers” in their performance. All last products they brought out are way better than Bitmains counterparts and totally outperform them … And what are the numbes for Bitmain and Baikal on Decred?

that Innosilicon asic are crazy, when can we expect the difficulty to exponentially rise?

Probably when the customers get them.

Bitmain and Innosilicon could turn their ASICs on and quadruple the difficulty overnight, but then no one would want to buy them for $10,000.

Their business model is very bad for their customers because the customers lack information about what the network hashrate will be after everyone else gets ASICs, while the ASIC manufactures have a very good idea.

3 Likes

yaaay zcash network entering new horizons

when zec price below 0.03 btc and diff 10M+++

imagine what diff would be if nicehash were into equihash…

happy mining

1 Like

…. wait longer on changing algo so U will see what happend last 51% atacss show what can asic mining do with others equihush coin when they fork and change algo no 51% attack

Do you ever mine cryptocurrency internally before you release a new miner onto the market?

No, we don’t do that. As in, sell a second-handed mining rig? If we sell a second-handed mining rig, we would be transparent that this is second-handed. We don’t do mining first and then make the mining rigs old and then sell them as if they’re new. If it was second-handed, we would tell the customer it’s second-handed.
It’s against our principles inside our company to do such things. This is absolutely bad rumors about us. From day one, we are transparent. We have ways to sell used rigs on the market. We don’t have to pretend that some mining rigs are unused ones.

And Billions don’t lie!

Not even used—I mean, would you ever, before you release a new miner, have a few months where you have an effective monopoly on that hardware, because it hasn’t been released publicly yet?*

No, it never happened. We have a small-scale test. We don’t do that. That is not our strategy.

I ask because there have been reports and rumors.*

I read those stories as well. There’s some kind of group of people that controlled the majority of the hash rate of Monero for a long time. I would like to say that to develop such a kind of ASIC [specialized computer chip] is not a kind of secret skill that only Bitmain has. Lots of people can do that. For large companies like Bitmain, and especially myself, I do not even have the time, attention or resources to have such a kind of plan to do this. If we develop hardware, we just release and sell it on the market. Right after we have sample machines working, we start sales to market. We don’t have such kinds of advantages.

As he said, it would be hard to hide that sort of thing: miners would notice dust and it would be more likely for the hardware to fail within the warranty period (bad for business).

Difficulty is going to go a lot higher!

I think this is proof positive 50Ksol/s 620W, Innosilicon announces World Best Equihash Miner A9 ZMaster that Zcash Equihash POW as failed, in any other industry this would be assessed as a security breach, here we have @Zooko spinning it as not, it is weird.

1 Like

It’s not that Equihash has failed, but the parameters chosen by the Zcash developers weren’t as memory hard as they expected. A coin called Zero used Equihash with different parameters that make it more memory hard and more resistant to ASICs. The current Equihash ASICs will not work on Zero and it’s doubtful that ASICs developed for Zero would be much more efficient than GPUs.

What I don’t understand is why there seems to be any resistance from the Zcash developers to changing the Equihash parameters to make it more memory hard. This could be rolled out in a few weeks or months.

(post withdrawn by author, will be automatically deleted in 24 hours unless flagged)