Topic: Crazy pow power reduction idea (Read 492 times)

Actually now with the mempool run-on the bigger threat is governments (not just any government, that Chinese mining province whatever-its-called in particular) ordering mining farms in their countries to shut down or otherwise cut off their power. Low fee transactions will eventually hit the 200MB mempool limit on most nodes and be dropped, and if this continues for more days then we could even see moderate-fee transactions hitting the wall too and this could be very disruptive until the next epoch.

1 sat/vbyte transactions already exceeded this cap. 60-70sats transactions are currently at 32MB according to johoe and coal plants are showing no signs of signaling an "all clear restore electricity now".

Which is generated by mining gear that can be bought with $.
If a government would simply print $ and offer 100k per s19pro they would be buying all the hashrate required in a matter of days.

odolvlobo is right.
The security is generated from the cost running this and all of what involved in it can be quantified and paid for in fiat.
The hashrate is an expression of that said cost, not a clear indicator of it, a 110th hashrate can be achieved with $10k and 5$ per day electricity consumption it doesn't mean that the network 5 years ago could be attacked with 5$.

As for the threat of the government, look at what's happening in China.
One order and all the mining is shut down, what could stop them from actually seizing those? Nothing!
Just because they haven't tried yet and in my opinion, they will never do it doesn't mean they aren't able to.

If people understood this and that the value created can be channelled into solar/windmills power consumption would be a topic of how btc saved the world from polluting coal/oil

If people understood the subsidies given to oil and coal and not given to solar and wind they would freak.  POW mining could create 3 watts of energy for every watt it burns.

I know this because we built and paid off a large solar array with BTC profits.

When I think of what is allowed with coal and oil and not done for solar and wind it is so freaking annoying.

See, I have to disagree with this assertion..
If it only took dollars $$ to mine blocks, then the US government and/or central banks could 51% attack Bitcoin at any time easily..
They have all the $$ dollars in the world, they could just print them, they have all the electricity they would ever need at their disposal..
The only thing they DO NOT have is access to enough computing power to come up with the needed hash rate..

Therefore IMO hashrate alone is what keeps Bitcoin secure..


BTW the US government has all sorts of nuclear reactor power plants that are military only and completely independent from the civilian power grid that can make ridiculous amounts of electricity..
Aside from ones at their bases and classifies ones, every modern warship and sub has a massive nuclear reactor they could use to power anything imaginable..

“measured by the cost of mining a block.”
I think this is a fine thing for energy consumption apologists to say, but isn’t very true..
Cost is no object when it comes to central banks and governments..
More like “security is measured by the difficulty of pulling off a 51% attack”, and the most difficult part would be amassing the computing power, not the money, nor the electricity..

Refute this..

It is not a good thing or a bad thing. It is just reality. Increasing efficiency is ineffective if your goal is to decrease energy consumption. Also, the security is not measured by the hash rate. It is measured by the cost of mining a block.

I fail so see how this is a bad thing..
Greatly increasing the security of Bitcoin at the same electrical cost sounds great to me..

HAHA I meant a tiny country. The power consumption is a lot and it is for a reason, the high security of a global decentralized money doesn't come cheap.

It appears that sometimes I terribly suck at math

Then it would starve the rest of the world and the accumulated demand means either the manufacturer leaves the country or other competitors in other jurisdictions come along to fill the gap. This also raises red flags for bitcoin community which would start planning for the possible attack before it happens (while the government is accumulating ASICs).

I hope I am not repeating what others have already written...

Your proposal is ineffective.

Regardless of how you structure it, PoW is designed such that the cost of mining approaches the value of the block reward. So, even if you only have 2 minutes of mining, a block reward's worth of energy will be expended in those two minutes.

Your proposal is similar to proposals that attempt to get useful work out of mining (such as generating useful information). That also does not work in general because the money gained by using or selling the PoW is spent on more hash rate.

Similarly, improving the efficiency of mining equipment is also ineffective because the money saved by efficiency is spent on increasing the hash rate. If a machine uses half the amount of electricity, the miner will run twice as many machines.

It costs about $3.90 per S9Pro, not $390k, lol. Each S9Pro consumes about 3250 watts per hour, or 78,000 watts per 24 hour day. To convert watts to kwh, you need to divide the number of watts by 1000, not multiply by 1000. The cost to power 1.5mm S9Pros would be about $5.85 million per day at 5 cents per kwh.

I think you are overestimating the cost of HVAC units. A HVAC unit has a very long useful life, and can be used for things other than cooling ASICs. In the past, I have also mined with ASICs and did so without having the air conditioning on due to the outdoor climate.

A government could also use the ASICs in a facility that is cold such as Northern China or Alaska. 

Wanna bet? If manufacturing is done in a country, the government gets to say what leaves the country. With manufacturing of ASICs done mostly in China, the CCP can force manufacturers to produce many miners, and prevent them from leaving the country via export controls. It could also seize the miners from the manufacturer.

You've missed this part:

Wait, are you saying that bitcoin mining is energy-intensive and rivals a whole country activity, of people eating, going to work, having fun, watching tv, and making transactions?  Infidel!


An s19pro burns around 90kwh of power a day, at 5 cents that are 4.5$, for the global network would be close to 6 million a day.
Not to mention that with 20% of the hashrate you would be making $10millions in profits from mining a day, a fact you completely ignored.


No, we're not!


P.S.
You're again ignoring the numbers in my scenario and it's a bit...you know..


10 years? More like 4 that counts and probably that one split in half due to faulty equipment and low hashing power of the older batches.
In 2014 (7 years ago) Bitmain released the S3, it was doing just half of TH/s, at that time just 1000 modern miners would have been able to launch a 300% attack

Your numbers are a bit off. Google tells me that s19pro price starts from $17,799 and each can produce 110 Th/s that means to get ~160 Eh/s you need 1,454,545 which costs $25.88 billion not 2.

You are also forgetting that just owning ASICs is not enough, they have to be run and that requires a tremendous amount of electricity. So basically the malicious government has to shut down their country (or a big part of it at least) to mine bitcoin with their 1.45 million ASICs! Not to mention that the cost of this electricity is also added on top of the cost of the ASICs.
That is about $390k daily per ASIC which $560 billion for all 1.45 million ASICs per day for 5 cent per kwh.

There is also cooling costs since you can't just let 1.45 million ASICs just run without cooling them or they'd blow up. The cost of that is also in the billions.
Now are are suddenly closer to trillion territory.

There is also another problem that the manufacturer can not suddenly produce 1.45 million ASICs and give ALL to this malicious government. It is also starve the market, you suddenly see globally that no new ASICs are coming to miners.

P.S. Getting current hasharate is 160 Eh/s to have 50% of the hashrate you have to have 160 Eh/s

The money may not be the issue but the logistics for manufacturing an equivalent of 10 years output of all the miner manufacturers would still be significant. If a Govt body ever attempted this, it would be hard to keep it a secret. Government agencies have gotten away with a lot in the earlier era when they performed nuclear tests and diverted the attention of a naïve citizenry from the ill-effects. I don't think it can be done in today's world.

With the benefits that Bitcoin and all of its spawns have given to millions of people worldwide, such a move to cripple Bitcoin would be hugely unpopular. Any potential hash-war between people and governments is basically a matter of perception. If they convince enough people that Bitcoin power consumption is a bad thing and that their centralized coins are better, we may have a problem at hand.

Those costs are overestimated, I've run in this discussion a lot of time.


Updating this, the price for miner stays the same as that's the cost to manufacture it, not what the current market pays for it, to make 20% with a  current 160exa you would need about 300 000 S19pro, that would be just around 750 million, powering them would require just about half of a modern nuclear powerplant full capacity.

And at this point, they would still mine coins for themselves, if people are getting profits at 10cents/kwh imagine how much they would be able to recover from the initial costs at 2-3 cents what power before taxes costs.


It would be more like the Anglo-Zanzibar War, fortunately enough no government has such plans, and hopefully, with the chip crisis going away the numbers in our favor will start to increase and with adoption people wanting to support the coin value and security might turn to non-profit mining, just for support.

The cost for trying to get that much hashrate would be so high that they ironically would have to print even more money just to do it, which hurts their own currencies and banks. [basically any commercial pool that attempts to cooperate with a government to do this will be dismantled by outraged miners.]

It would be some variant of mutually-assured-destruction from the Cold War era where the government's currency is crippled along with bitcoin.

If this is the US we're talking about, this'll also take down all the other currencies (at least temporarily) because of its status as the world reserve currency.

Dear author, do not buy Bitcoin energy consuption FUD distributed by deceptive PoS coins founders, their ignorant followers, and representatives of governments and institutions that feel threatened by a truly decentralized coin.

Who fail to mention that securing a monetary network with rent is just as energy intensive, ultimately. Because here, energy consumption can be postponed considerably, but it is just as inevitable (unless you believe that all capital owners are vegans who live in trees and never consume their profits).

If the level of Bitcoin security is to be maintained, the only thing you will achieve will be to increase the investment of miners in hardware and reduce their electricity bills, while the global economy will continue to use energy to produce this hardware (not to mention what a burden on the environment would be mountains of toxic waste resulting from the production of unnecessary additional mining units).

Now that the "green" energy revolution is underway and there are already countries producing electricity entirely from renewable sources, energy production and consumption does not always burden the environment.
Bitcoin mining overwhelmingly relies on the use of renewable energy sources (reducing renewable curtailment) and the use of stranded energy.

Bitcoin mining may ultimately accelerate the global transition to renewable energy and thus indirectly save the Earth and everything that lives on it.

Only applies to PoW-based cryptocurrency which uses SHA-256. If they plan to destroy another PoW-based cryptocurrency, they need to invest on different type of ASIC.


Good point, but also leads people to work on more secure alternative.


Or currently thinking solution that actually works on reality.

If they could outhash and destroy Bitcoin any other crypto would be a walk in the park to destroy next..
It would also destroy the confidence in all/any other crypto..

You don’t think the government would go to extreme lengths to secure their monopoly on printing/issuing money, with the full backing of banks?

I see hardly anyone is interested in protecting Bitcoin from governments anymore, but rather keeping it in compliance with regulations..
When on-chain KYC?

When laws become tyrannical the “good guys” become criminals..

What does that matter?
We may have to become criminals to use and sustain Bitcoin.. Good thing the entire point of Bitcoin is to be unstoppable and unregulatable no matter what silly laws they make..

Being a “good guy” has little to nothing to do with obeying laws..

The whole point of Bitcoin is to take power away from governments..


In 2013 we didn’t have the powerful ASICS we do now.. Neither did governments/attackers..
Our miners need to be able to beat the next top 10 most powerful governments combined to keep Bitcoin secure..
That’s why it’s not about electricity, it’s about hashrate..

Incredibly? Don't think so.
Just raw numbers in terms of hashrate tell only a bit of the story.
For example before the 2013 jump the total hash rate was well below what a s19pro can now deliver, was the network successfully attacked, or was it considered safe even despite the fact it had less than 1/1mil of the hashrate we have now?
The key behind those 160 and 800 in those two scenarios is the price required to acquire them, if this method is a flop and miners quit because of mining gears failure, contracts with powerplants getting suspended as drawing 10MW every 8 minutes from the grid are nothing any company wants to deal with the whole picture would change dramatically.

In case of a government attack, it will be the same stuff, how much it will cost them to do it versus how much can miners spend to protect the network, in the end, the numbers don't change. Besides, if I were in the place of an official tasked with crippling bitcoin I wouldn't go for a 51% attack, even a 20% of the hash rate used to mine empty blocks would trigger higher fees, enough panic to make the price go down, more miners out of business, more cheap second ger gear to acquire with the $ coming from mined and dumpedd coins, more empty blocks, rinse and repeat.


One thing I've earned about regulations and laws is that when you start making things complicated with hundreds of them it's the bad guys that manage to avoid them and get the upper hand.
Same here, one guy finding a weakness, and the "good" guys are screwed.
 

Nodes do not normally "reach a consensus" on anything. Nodes will validate each block according to consensus rules, confirming that each block is valid, and that no blockchain exists that has a higher total POW.

No matter what, someone will have the ability to know the "random number" ahead of time, and could use this information to their advantage, either by mining personally, or selling the information to a select group of miners. It is possible to prove that you knew information on or after a certain time, but it is not possible to  prove you did not know information prior to a certain time.

If the difficulty target block time is reduced, there will be more stale blocks because there is a higher probability that two blocks will be found in the time it takes for a found block to be propagated and validated by the network. This has nothing to do with a malicious actor.

In the case of competing blocks at the same height, the current status quo allows a "race" to be potentially resolved very quickly if a subsequent block is quickly found, however under this proposal, it will take a minimum of 8 minutes to resolve. 

Think manhattan project energy..
No, not to make 50btc or even 5,000,000 BTC.. Their goal would be to DESTROY BTC..


Yes, miners would eventually get back to the limits of their efficiency in their competition..
If BTC was running at 800exa rather than 160exa, it would be incredibly more secure against government attack, would it not?

Yes mining decentralization of the hash is key to its security, but the amount of hash in the hands of the “good guys” matters too..
The more hash running on Bitcoin in decentralized hands of good guys, the more secure it is against government attacks..


I don’t know..
Do you have any ideas?

No, it will not.
Power consumption is capped by the reward, miners will never be able to spend more on energy than what they make from the reward.
Even if your plan would work things will simply change, instead of constantly deploying 160Exahash of hashing power miners will afford to deploy 800Exa for 2 minutes, in the end consuming the same amount of power and getting the same reward.

This is just like fishing quotas, when you restricted the number of days boats could fish there was simply a rush in those days and it ended with more boats getting launched ending with a bigger catch per day overall so the effect was zero, thus it changed to quotas. But unfortunately, there will be no way of doing this with bitcoin as there is no central authority issuing permits to mine.

But that energy is not available.  It would cost a blackout in significant share the country. And the united states is not a dictatorship, but a democracy.

First of all, nobody would allow such a blackout for a day in a democracy which would have very severe consequences(people dying in hospitals for example and billions of usd lost in the economy, )

Even if that could happen, There are responsibilities,  accountability  and other tools which certainly would be used against someone who was responsible for a blackout like that. Certainly many powerful people would he removed from the administration . For what, 50btc?

How's that true? What is the power consumption required to mine that it is reduced? You, yourself, stated it that miners will be able to spend that energy, they'd spend within these 8 minutes, on more ASICS.

But if they spent more money on ASICS, then the attacker(s) could do it too. It wouldn't change anything with the network's security.

I'm trying to understand if you do understand how mining works. I like that you proposed this idea and I hope we continue this discussion further, in order to convince some people, but the answer to this question is:  It depends. If the mining income is greater than the electricity expenses, then mining is more profitable. As long as mining is profitable, more miners will join to earn money out of this procedure. As long as more miners work on extending the chain, the difficulty increases, which makes mining less profitable. So it wouldn't increase the profitability of mining, neither paying less for electricity.

If the old inefficient miners "skyrocketed", then what about the new efficient ones? Note that this could be true only for the short term. On the long term, they'd stop being profitable, as they're now.

You shouldn't be confusing hashrate with security. The more the hashrate doesn't mean the more the security. The security of the network is assured by the hashrate's distribution. If the hashrate is 100x than now, but 70% of it is owned by 4-5 pools, then it isn't as secure as it was back when you could mine a block with your CPU. And that's because the hashrate was distributed. The more the hashrate means that it's more difficult for an outward to perform 51% attack, without going through a pool.

---

Do you know what's the problem with mining? Man's greed. You shouldn't be looking how to reduce the power consumption, with a different PoW mechanism. This is a problem that existed long ago, before bitcoin. To me, when greed is uncontrolled and it's damaging, then that's a big issue. Fortunately, greed with bitcoin mining isn't uncontrolled. Some governments can illegalize the use of ASICS, which would help with the power reduction and wouldn't change the security of the network, neither the number of bitcoins mined every day, in contrast with gold mining.

Sometimes a block isn't even found in 10 minutes, it's found in 4, or 25, or 60. You get the idea.

For a long time nobody will be able to find a block because the time limit is shorter than the time it takes to continuously hash SHA256 to find a solution. Then in the next epoch whoever gets their miners started the quickest can mine whatever they want in that timeframe (including invalid transactions and wipeout of others).

I do think that the idea *could* be possible but the problem is with reaching the consensus for the variables that you've added. The network's time can deviate between node to node by quite a bit and it is difficult to synchronize all of them without a reference to a central server. Transactions aren't relayed instantaneously and there can be a delay between the nodes seeing it.

As for the benefits, I fail to see how this would increase the security at all. Given that you don't have to mine for 8 minutes, this makes it such that all of the work is done within the 2 minutes which would be much lower than the 10 minutes of continuous hashing that we have right now. It wouldn't bring older ASICs back into the game as the profitability would remain roughly constant; think of it as Bitcoin having to decrease the difficulty retarget to make 2 minute blocks. Standby power could still be fairly massive and would be much more inefficient than continuous hashing.

That's right, it was poor choice of words. What happens is an increased number of stale blocks whenever difficulty is reduced and if there are malicious/competitive miners active on the chain. The effects are similar to a 51% attack since the last block that was mined is going to be replaced over and over.

Because block time is the result of the difficulty value not a standalone variable.
To reduce the block time we have to reduce the difficulty, and when we reduce difficulty a miner (or a mining pool) can find next block easier. And when the reduction is 80% the mining pool that has a moderately high hashrate may even find the next block at reduced difficulty instantly. That means 2 competing mining pools with high hashrates could end up competing with each other and reverse the other pools block multiple times. This happens because the cost of performing the attack was also reduced by 80%.

The problem is coming up with that random number. Your "using transaction" idea won't work. But maybe we could add more complexity to this hypothetical scenario where miners have to "mine" something else. That is do another sort of work to find something that takes 8 min to finish and can't be done on their SHA256 ASICs.
It could be another PoW hash using a different algorithm or some mathematical problem like finding next prime gap of maximum known merit (GAP coin!).

I would contend that the security comes from sheer computing power/hashrate rather than how much energy is used to run the computing power..

It’s “the next 500 supercomputers in the world can’t compete”, not “the next 500 power plants in the world can’t compete”..
(Probably much more than 500 since this saying was a thing, and it’s never been about how much electricity it would take)

If the NSA wanted to 51 Bitcoin, do you really think they would care how much electricity it took?
Neh, they can make plenty of energy.. They have the energy, what they don’t have is the hashrate/processing power..

It would only take about 1 day to 51 Bitcoin, and they could use an entire nuclear power plant, or 10..
The cost of burning that energy for one day is nothing compared to the cost of burning that energy 24/7 to KEEP it secure..

Energy doesn’t matter, it’s the hashrate that keeps Bitcoin secure, because they don’t have the hashrate to compete no matter how much energy they have/use..

Bitcoin could keep the same hashrate and use less energy, by only hashing in bursts, thus allowing miners to buy even MORE hashrate with their savings, and even put all of the old currently inefficient miners hashrate back online to increase hashrate, because it would be profitable again..
Might even make CPU mining profitable again until hashrate grew to make it a competition of efficiency again..

Would incentivize a massive amount more hashrate and INCREASE DECENTRALIZATION to higher energy cost areas..
Mining on almost anything would become profitable until hashrate atleast tripled..
Hashrate would go through the roof.. And thus also security..


If you want blocks/solutions found in 2 minutes you adjust difficulty so on average a block/solution would be found in 2 minutes with full current hashrate right?
It’s hashrate vs hashrate..

I don’t see how reducing the block time to find a hash faster would somehow give a minority hashrate(pool/attacker) any advantage.. They would still need 51%..


Ok, average of node clocks having passed 8 minutes of idle time before the 2 minute hashing session begins (with difficulty set to take an average of 2 minutes for X hashrate to solve a block solution)

I think you would still need a random number published after the 8 minute idle time hashed into the next block to PROVE that the miners only starter hashing after the 8 minute idle time is up, or they would just mine the entire time and hold the solved block until after the 8 minute idle time, in which a cheater could 51..

The point of the random is to MAKE them wait the 8 minutes before they started hashing, or they would be hashing for nothing, because without including the random number (starting pistol shot), their block would be invalid.

I’m still talking 10 minute block times, but instead of hashing for all 10 minutes of every block, a random number could be released after 8 minutes, like a starting pistol shot, for a 2 minute hashing sprint at the end of every 10 minutes..
Their solution would have to include the random to prove they started after the 8 minute idle..

8 minute idle-release random after the 8-2 minute hashing sprint (difficulty adjusted to find solution in 2 minutes)-block still found/added in 10 minute increments

Would probably also reduce the variation in actual seen blocktime too..
Variation in shorter block times is less (I imagine it’s likely a percentage), and would still only have blocks every 10 minutes on average to keep the blockchain a manageable size (decentralization)..



I do find this to be a valid point..

I’m not a code genius..
How could a starting pistol shot random number be programmed to be broadcast after an about 8 minute idle period to start the 2 minute hashing sprint?
Thus proving they didn’t start hashing until the idle period was over..

The way I’m imagining it, until this random number was broadcast after the idle period, they wouldn’t even have anything to work/hash on, because they don’t yet have all of the information they need to create a valid block at all..

Nodes would reach a consensus as to what this random number is after the idle break, and not accept a block that did not include it the random number.. Block would be invalid without it..
Somehow..

This wont create an opportunity for a 51% attack. The ability to pull off a 51% attack is based on the attackers hashrate as a percentage of the total network hashrate, it is not a function of the difficulty.

You don't need to generate random numbers or anything crazy like that, it can be done by changing how next target is computed by using different number of previous blocks to get the time. For example "If last block was mined in 2 minutes, reduce difficulty by 80%; Else set difficulty as it should be based on past 2016 blocks". Roughly similar to what we are doing in testnet but only when the time between blocks are 20 minutes.

Here are two major problems with the idea:
1. As it was mentioned above, there is no universal clock in the decentralized network that determines "what time it is". Your node may say it is 14:05, my node says it is 14:30 and another can say it is 15:00. That's why we take the average of all times we get from the peers.
2. When you reduce the difficulty, the hashrate doesn't go away. It is still there and when the difficulty is reduced by 80% that massive hashrate can find the next block in an instance. Now we are facing an easy to perform 51% attack.

PoW power consumption is what makes bitcoin secure. IMO there is no problem to solve here. There is big propaganda made by altcoins criticizing bitcoin TPS and energy consumption...

You are basically staking energy, which is something from outside the blockchain, something valuable in our world and society as we know it. This is one thing that gives value to bitcoin. (the opposite of staking coins, which you are staking something from within the blockchain itself)

Bitcoin is not wasting energy to generate blocks, it is using energy to create value and utility.

Miners also mine in places where energy is cheap, and those places are usually where the generation of eletric energy is bigger than the consumption.

I saw just now this video from Antonopoulos where he explains that it is easy to bash bitcoin energy consumption, because it is transparent. But you cannot see how much energy banks consome (with buildings, bureaucracy, security, guards, etc)

This won't work.

Who determines what the random number is? Transactions are not timestamped when they are broadcast, and there is nothing to stop a miner from broadcasting their own transaction 8 minutes after the prior block was found.

Also, block timestamps are not the exact time a block was found, it can have up to a two hour variance, so it is possible for block n to have a later timestamp than block n + 1.
It takes time to shut down a miner, and to start a miner. The power savings would not quite be 80% under perfect circumstances.

I think this would encourage miners to backdate (or back time) found blocks, and withhold the blocks until the next block is broadcast, creating many orphans.

Reduce difficulty to target block time of 2 minutes..
After block solved nodes start timer to 8 minutes..
At 8 minutes somehow generate random number  (possibly use last/nearest tx broadcast closest to 8 minute mark as the random number?)..
Next block must be solved including hash of last block and random number generated to prove work started after the 8 minute mark since last block solved..
Still requires the same amount of computing power giving same security, but miners only run for 2 minutes out of every 10 minutes instead of constantly..
Reduces power consumption by 80%..


Have any ideas like this came up to run POW security in bursts rather than constantly?

Could it possibly make BTC even MORE secure because miners would be able to spend that 80% power savings on more ASICS rather than spending it in electricity?
This would greatly increase the profitability of mining, not paying so much electricity, and create even more of a race for more hashpower, no?
Would probably even make mining profitable even at high electricity costs until 500% more hashrate came online to compete..
The value of miners, even old inefficient miners, would skyrocket? (Basically all miners would see an efficiency increase of 500%?)
Could insanely increase hashrate and hence security?