In the end, I think alle said here can come trough.
I have no idea where it will go. Not a big believer in Asics, I think in the end all small miners will be gone, also the small Asic miners. There will be a weaponsrace ( hashrace ) and the big ones will survive. I think in the end their will be a POS . Mining since Bitcoin, but since last month I am thinking more and more my mining will be done soon. I love running and tweaking my rigs, don’t like plug and play.
Supports what I’ve been saying all along: A slight tweak to Equihash would invalidate these ASICs.
The idea that ASICs can adapt to proof of work changes is a bunch of nonsense peddled by people who are trying to sell ASICs.
2 Likes
Proof of work tweaks are part of the arms race.
In fact, proof of work tweaks are an essential part of the arms race, because without proof of work tweaks, it’s far more likely that one group will amass the majority of the hashing power.
If a cryptocurrency is not committed to proof of work tweaks, then it should just use proof of stake.
2 Likes
That’s why I think PoS will be the last station.
I think Asics will be smarter and smarter and can be updated with every pow tweak. Could be wrong here, if so, please enlighten me ( and I really mean this in a nice way )
51% attacks problem more about renting hashrate, than increase net hashrate. Even now, there is not enough hashrate for rent to attack ZEC and never will be (excluding ASIC’s of course). It doesn’t matter from whom they (nicehash and others) rent computing power, GPU or ASICs. So basically if you add ASIC’s to the equation - nothing changes to ZEC security, you simply kick out GPU miners and replace them with ASIC miners. And all this under the guise of network security.Thanks.
3 Likes
Actually this is not really correct and i realized that my own calculation on 51% attack on gpu projects is totally wrong.
The calculation on the site https://www.crypto51.app/ “PoW 51% Attack Cost” is onlyt correct and valid for the Asic coins as the hashrate their for rent is fixed and can not be shifted.
Explaination: Last night i noticed that for a short time the Lyra Price went up as high as 13 USD/per 1080ti for a short period (mostly due another ongoing hack attempt on Verge). Immediatly more hashpoer was shifted from other algos to Lyra as it was more profitable to sell on lyra as on other algos (Nicehash, similars).
After following this i realized that while the Asic hashpower there is a more or less fixed contstant it’s not the same with gpu hashpower. On whichever algo on the nicehash marketplace a buyer offers more $ for hashrate the more hashpower gets shifted there. Just logical when someone thinks about it as the main intention from selling hashpower to Nicehash is to make as much as possible money, no matter what algo as the more profitable nicehash is the more hashpower it attracts of course. There are even facebook and other other groups that share what algo on nicehash is the most profitable at the moment and when to switch there, just as a side note.
This having in mind, analyzing over several days the site https://www.crypto51.app/ that calculates the 51% attack cost i came to the conclusion that it takes only in account the immediate and free hashpower of a given algo but NOT calculating the possible and likely shifting hashpower.
Someone can make a conversation list for the different algos and free for purchase hashrate on the different free gpu algos and convert them to any other given algo to see how much the theoritically rentable hashpower would/could be in case someone wants to buy up as much hashrate as possible.
I didn’t have the time yet to add it up theoretically, but there is a good chance that only Ethereum would be on the safe side of a possible theoretical 51% attack attempt…
P.S.: This is based on theory and just finding a flaw in the https://www.crypto51.app/ list for the gpu coins listed there.
1 Like
Please update the calculations when you get the chance.
Maybe we need to add some nicehash-resistance in addition to ASIC-resistance. Actually, algos like X16S provide it by unstable hashrate from one block to another, which makes it difficult to resell such hashes.
Are ASIC’s truly necessary to prevent these attacks? Please correct me if I am wrong but a coin’s value to mine is attached to its’ hashrate. Replacing 500 Msol/s of GPU’s with ASIC’s is still worth the same to mine. So why would ASIC’s increase the hashrate and make the network safer? Can’t exchanges just increase the confirmations to around 100 to have time to verify if an attack is happening before someone can cash out or exchange the ill gotten coins? Won’t that put an end to the 51% attacks? I believe you are implying that ASIC miners are willing to make less money and/or miner’s with larger operations like ASIC’s more so they can afford to make less per ASIC. Otherwise, I don’t see why ASIC’s make the network safer. ASIC’s are rentable as well.
1 Like
This looks interesting.
1 Like
there are pro and contras for that. I several times my own thought about that to be honest. While most are fixed on Asics i realized, especially after seeing the https://www.crypto51.app/ list that one of the biggest dangers for mid to small projects actually are coming exactly form services like Nicehash …
Pro:
A strong argument for Hashrate rental service restriction would be of course reducing the chance of 51% attacks, no doubt on this
Contra:
- A lot of people would e driven out of mining that don’t have the technical ability to manage things their own.
- Nicehash is a perfect place to begin with mining, i admit i begun there as well than back the first weeks until i got more advanced and figured out more how to make things better.
- in normal case and without bad actors these services are indeed good as a lot of real small miners are there. I know even of projects that buy hashpower for their own projects there to maintain their network.
This said i’am not sure if such restrictions would be ok. Eventually some more regulated mining hashrate distribution would be better. I mean right now there is no verification, no nothing. You send some BTC there and buy as much hashpower as you want for whatever you want. I think this is eventually something services like nicehash should begin with, some kind of self regulation.
If I were nicehash, I would limit the amount any given entity or network could be rented and pointed anywhere.
If they would care they could actually do that and much more, but than again. The more hashrate they sell the more the hashrate sellers get. The less they sell, the more unrented hashrate is on the net reducing the sell/buy price. So they i doubt they are currently really interested in restricting whatever …
But i agree, something should and could be done, be it KYC and not just anonymious selling be it limits or restrictions…
In theory everything is possible, but we live in real world, where you can’t get enough hashrate to 51% attack ZEC. And when ASIC’s replace GPU all (net hashrate, hashrate for rent) will be same x12/x50/x100, excluding you profits.
It sounds indeed interesting, and it’s maybe a niche he found but it raises again some general concerns imo:
-
Will these units be good enough to competete with later competitors or is it just a some time a given projects safes until other Asic producers come and could make better more Asic efficient miners?
-
I doubt he is doing that for free. Pretty sure this will cost the projects a good sum of money. I doubt obelisk will invest some 10M$ for R&D just in good faith and trust…
-
While in theory and on first read it sounded good on second read i immediatly thought about what if the devs/company just go hand in hand with obelisk? From not selling all hardware to the community up to having obelisk stealth mining…
Just some thoughts, but why not, we will see how such projects work…
Do you mind sharing the calculation? I guess you made one after you came to that conclusion…
Mining Rig Rentals | Equihash/zcash Mining Rigs - 36.2 Mh\s
NiceHash - Leading Cryptocurrency Platform for Mining and Trading - 46.8 Mh\s
some others - 20 Mh\s
Total - 103 Mh\s
Prove me wrong and find 250 Mh\s to rent and not in theory, just from availability
Just call Bitmain and you are set 
2 Likes
1st part:
Nicehash, free rent power at right now 07.06.2018
First number EU available, second number US server available
Keccak: 2.5752 TH/s 4.5847 TH/s
Neoscrypt: 64.6683 GH/s 65.6603 GH/s
Lyra2REv2: 6.233 TH/s 6.4431 TH/s
Ethash: 2533.9996 GH/s 1619.1668 GH/s
Equihash: 27.1378 MSol/s 19.5731 MSol/s
Skunk: 6.3027 GH/s 4.6646 GH/s
CryptniteV7 68.1828 MH/s 50.2889 MH/s
Cryptonite Heavy: 1.791 MH/s 1.6901 MH/s
Actually it proofes me right. Some days ago when i made another calculation only for equihsh their have been around 75 MH/s available, now we see it’s about 30 MH/s less, means these hashrates gets shifted heavyly on the market place on wherever there is demand:
Someone to make a conversation what this would be in equihash only so the calculation is more decentralized? :-)))
You raise some interesting questions, please copy to the following thread as they are relevant:
Ballot Proposal: Embracing Simple PoW and ASICs by acityinohio · Pull Request #21 · ZcashFoundation/Elections · GitHub
Note that Obelisk is the company which launched Blake2b miners for Siacoin but were beaten to market by Bitmain and others.
2 Likes