Mining Centralization Concerns

[RealBitcoin]

replying to you and continuing point 4 of 'realbitcoins' post

4. the heat coming off a chip will burn out the batter of a iphone/android.. if not ur hand first. putting chips into any piece of home appliance is meaningless.. if there are only 20 pools. it does not matter of there are a million users.. because a manufacturer will probably have more hash power alone compared to the 50k people per pool

I answered this in the previous post

we are no longer in the solo mining era where competition is measured by individual users/equipment. its done by the combined power around the world linked to a particular pool server. (emphasis on the pool server being the main thing)

Sure but we can do anything to make the process as decentralized as possible.

its also worth noting that asic chips have NO HARDDRIVE.

Indeed this is an issue now. Maybe we need some remote mining BIPs in the future so that miners could mine without being nodes.

Or lightweight mining where miners would only need the past x blocks to validate stuff from.

[chek2fire]

i fell more safe that most of the miners are in china than to be in USA. USA is a very corrupt country with agent organizations like NSA, CIA to act in the shadows and without any democracy permission. Of course China has no democracy but this paramilitary organizations as i say in USA is a danger to freedom.

[RealBitcoin]

i fell more safe that most of the miners are in china than to be in USA. USA is a very corrupt country...

China is communist.

And they have totalitarian internet censorship.

I would not put my money on China.

[franky1]

2) Not necessarly that way, but what I mean is that the manufacturer to sell the parts to assembly plants. The manufactoring factories have always the lowers profit margin so, an extra middleman here is actually beneficial. Then if you worry about assembly plants being greedy, then just have many of them. It's easier to assemble it from parts than to make the parts, and it's the assembler that would be the greedy one. So just have more assemblers than manufacturers and the price should come down. An extra middleman in this case is probably good.

that worked out so well for Butterfly labs.. 2 years on and they couldnt even release an end product.

its actually cheaper to have the manufacturing and mining facility right next door to each other (1 hour delivery at most). instead of making the chips in china/europe shipping them to america and then have someone at $9 an hour making the rigs and then another group of guys at $9 an hour supervising the mining(fully produced running asics) facility.
sorry but america wont EVER be a good production centre for asics

No because then they just mop up the profits, because they create everything, they will entitle themselves to 8x profit margin as no competitor exists that can compete with the complexity of their business.

If they just make the parts, then the profit margin normalizes to 10-30%.

And the assembler just charges based on supply and demand. Then many hobbyists can just turn this into a business, they buy the parts at factory rates, and assemble it in a garage and sell the mining equipment on e-bay or openbazaar

So the overall retail price then will decrease massively as the manufacturing process will be decentralized as well.

i dare you to buy a DIY car 'kit' and build your own car. and sell it.
i guarantee you it wont be cheaper then buying the car from the manufacturer. and i guarantee you the employee's of the manufacturer get it even cheaper again.

the only way to be competitive is if you follow bitmains business plan and do EVERYTHING inhouse. not only will you save on the shipping and labour, but the delay in times of shipping, fabricating and then ultimately selling..

so there is no point making chips in china/europe. ship them to america and some garage geek makes the rigs in america to sell on open bazaar.. it should wont be price or time friendly in regards to mining rewards

[chek2fire]

i fell more safe that most of the miners are in china than to be in USA. USA is a very corrupt country...

China is communist.

And they have totalitarian internet censorship.

I would not put my money on China.

as i say if i have to choose between the "capital-communist-Frankenstein" China and Usa i will definitely choose China. I dont have trust to USA because they have too many organizations that they do whatever they want without any permission from anyone.

[franky1]

i fell more safe that most of the miners are in china than to be in USA. USA is a very corrupt country...

China is communist.

And they have totalitarian internet censorship.

I would not put my money on China.

lol im guessing you watch fox new's interpretation of communism.

[RealBitcoin]

i dare you to buy a DIY car 'kit' and build your own car. and sell it.
i guarantee you it wont be cheaper then buying the car from the manufacturer. and i guarantee you the employee's of the manufacturer get it even cheaper again.

the only way to be competitive is if you follow bitmains business plan and do EVERYTHING inhouse. not only will you save on the shipping and labour, but the delay in times of shipping, fabricating and then ultimately selling..

so there is no point making chips in china/europe. ship them to america and some garage geek makes the rigs in america to sell on open bazaar.. it should wont be price or time friendly in regards to mining rewards

Ok have it your way, the no middleman version is frictionless and should be cheaper, but I tell you that in high-tech manufacturing you wont find a big supply of things.

Aside from Silicon Valley, China and Israel and Japan, there arent many high tech manufacturing countries left, so you can probably count the total manufacturers with your fingers.

Or they really need to convince everyone in the industry to stop making smart phones and start making bitcoin miners, which will be hard

I`m not sure how many idle factories you will find for this.

[Soros Shorts]

The concept of millions of miners mining small amounts at a loss has some merit. I wouldn't mind plugging something into the wall that costs $5/month in electricity and generates $1/month worth of BTC if millions of other people were doing it in order to keep mining distributed. No need for a fancy chip in my phone or car. Devices like the AntRouter but one that can mine on any pool would be ideal.

[chek2fire]

The concept of millions of miners mining small amounts at a loss has some merit. I wouldn't mind plugging something into the wall that costs $5/month in electricity and generates $1/month worth of BTC if millions of other people were doing it in order to keep mining distributed. No need for a fancy chip in my phone or car. Devices like the AntRouter but one that can mine on any pool would be ideal.

i think this is the future of bitcoin mining. The mining era that we have now is near the ending. Imo we have 4 more years with big miining data centers and after that the mining definitely will pass to lot

[franky1]

The concept of millions of miners mining small amounts at a loss has some merit. I wouldn't mind plugging something into the wall that costs $5/month in electricity and generates $1/month worth of BTC if millions of other people were doing it in order to keep mining distributed. No need for a fancy chip in my phone or car. Devices like the AntRouter but one that can mine on any pool would be ideal.

USB block eruptors already exist. yet the issue is not the distribution of the asics. they do not store the blockchain after all... no asic does..

its the POOLS that need distributing.

and this will only happen if there was enough variety of getting asics from different locations at a price that competes against the main players.
otherwise people are just going to contract to an ASIC that is maintained remotely by couple the manufacturing companies (like they do now)

a good way to force it. is to get nodes to ignore a block if it was created by the same pool in the last X amount of blocks.. this would equal the chances of many pools getting a chance and slow down the dominance of just a couple.

getting a rig delivered to your house is not profitable nor helpful. but expanding the choice of pool companies that also manufacture is.
if there was atleast 10 manufacturers of asics that house the rigs for users to contract to remotely, then there are more chances and choices of pools. while letting the whole world have a chance

i know people have dream scenario's of the utopian idea of how mining should be. but that would only ever last 2 months before people start to change back to old ways due to profit. so we might aswell stick to realistic long term case scenarios of expanding the manufacturing and pools. as thats the important part.

individuals with a micro miner embedded in a phone is meaningless for a multitude of practical reasons

[AmDD]

I had this discussion with a friend not long ago and here is my thoughts... This came about when talking about 21INC and Intel being involved in blockchain tech.

There are about 2 billion PCs in the world at the moment. If each motherboard had a single ASIC chip from the AntMiner S7 (25GH/s @ 5 watts) it would equal 50billion GH/s. The entire network is about 1.3 billion GH/s at the moment. 8 watts isnt going to break the bank for any homeowner, or even a large corporation I wouldnt think. The miner could be configured in the BIOS and locked down somehow to help guard against botnets taking over systems or even a physical switch/jumper on the board that when opened would allow changing of the mining address but when closed was impossible to change via software. like old 3.5" floppy drives with a lock switch.

As mentioned above, the biggest issue will be where they mine to. If to a pool then its not as decentralized. Solo would be ideal but with only 25GH/s the chances of hitting a block would be insane and so most people wont bother to keep track of a wallet address and check it so the coins would be mined and lost. You would need some third party way to monitor the address's and notify users of a mined block.
or even handle distributing the coin but there again, thats not the point of solo mining, it might as well be a pool at that point.

Maybe you could incorporate a pool into a router. In todays world, nearly every home with internet access has a router. No one sets up a DNS or DHCP server in their homes but they are there. Why? Because its built into the router. This would mean they would have to add hard drives to routers which would increase cost as well as physical size, however I dont think it would be by too much. All motherboards (or any other ASIC enabled device in the home) would just point to the default gateway as a mining pool for solo mining. Each home becomes its own pool.

[lumeire]

the solution. expand the manufacturing and get more rig manufacturers on the market. which will trickle down as more choice, and more options and also a price war on the rigs. which will result in cheaper rigs to allow for more chance of competition.

Where is silicon valley folks? All those manufacturing plants...

Dont let them go to waste, they should start producing cheap mining equipment for americans and europeans.

Even if you produce cheap mining equipment elsewhere besides China, operating costs such as electricity and manpower would still kill your profits. In all honesty the only situation that can keep mining as decentralized as possible is for the difficulty to remain stable.

[Ultrafinery]

I had this discussion with a friend not long ago and here is my thoughts... This came about when talking about 21INC and Intel being involved in blockchain tech.

There are about 2 billion PCs in the world at the moment. If each motherboard had a single ASIC chip from the AntMiner S7 (25GH/s @ 5 watts) it would equal 50billion GH/s.

Are those boxen running 24/7? Never switched off? And Toto Laptops too?
Constantly connected to the internet?
Why would anyone pay an extra, let's say $25 (because arbitrary numbers are fun)  for a computer, knowing they will "mine" $5?
Is Grandma/Visa/my hospital interested in Bitcoin mining?

The entire network is about 1.3 billion GH/s at the moment. 8 watts isnt going to break the bank for any homeowner, or even a large corporation I wouldnt think.

When I was a kid, I reasoned like that too: "Most people don't even pick up a penny when they drop it, they couldn't care less. If everyone in the world just sent me a penny, it "isnt going to break the bank," but I'd be a millionaire!"

[pedrog]

the solution. expand the manufacturing and get more rig manufacturers on the market. which will trickle down as more choice, and more options and also a price war on the rigs. which will result in cheaper rigs to allow for more chance of competition.

Where is silicon valley folks? All those manufacturing plants...

Dont let them go to waste, they should start producing cheap mining equipment for americans and europeans.

Even if you produce cheap mining equipment elsewhere besides China, operating costs such as electricity and manpower would still kill your profits. In all honesty the only situation that can keep mining as decentralized as possible is for the difficulty to remain stable.

That's not, and should not be, an option.

A possible solution would be other variables gaining importance for mining, not just cheap electricity, for example, if blocks got bigger good Internet connections would be crucial for success when mining and the playing field could be leveled and a bigger percentage of mining would be done outside of China.

[AmDD]

I had this discussion with a friend not long ago and here is my thoughts... This came about when talking about 21INC and Intel being involved in blockchain tech.

There are about 2 billion PCs in the world at the moment. If each motherboard had a single ASIC chip from the AntMiner S7 (25GH/s @ 5 watts) it would equal 50billion GH/s.

Are those boxen running 24/7? Never switched off? And Toto Laptops too?
Constantly connected to the internet?
Why would anyone pay an extra, let's say $25 (because arbitrary numbers are fun)  for a computer, knowing they will "mine" $5?
Is Grandma/Visa/my hospital interested in Bitcoin mining?

Do you pay extra for your computer because it has 2 extra USB ports? or even firewire? No, its so cheap for them to add it that its just added. You dont sell it as a mining device, you sell it as a PC with an extra feature of mining. Ive never once used a firewire device but I have owned lots of systems that have the feature. If its not something you are interested in then you dont set it up. This is the same idea as the 21inc computer. Those who view it as a mining device are missing the point.

The entire network is about 1.3 billion GH/s at the moment. 8 watts isnt going to break the bank for any homeowner, or even a large corporation I wouldnt think.

When I was a kid, I reasoned like that too: "Most people don't even pick up a penny when they drop it, they couldn't care less. If everyone in the world just sent me a penny, it "isnt going to break the bank," but I'd be a millionaire!"

Maybe this could be worded differently. Id be willing to 'donate' $0.50 a month to run the mining chip in order to help secure the network. Those who would not be can simply not configure the miner and it will sit idle.

[pedrog]

I had this discussion with a friend not long ago and here is my thoughts... This came about when talking about 21INC and Intel being involved in blockchain tech.

There are about 2 billion PCs in the world at the moment. If each motherboard had a single ASIC chip from the AntMiner S7 (25GH/s @ 5 watts) it would equal 50billion GH/s.

Are those boxen running 24/7? Never switched off? And Toto Laptops too?
Constantly connected to the internet?
Why would anyone pay an extra, let's say $25 (because arbitrary numbers are fun)  for a computer, knowing they will "mine" $5?
Is Grandma/Visa/my hospital interested in Bitcoin mining?

Do you pay extra for your computer because it has 2 extra USB ports? or even firewire? No, its so cheap for them to add it that its just added. You dont sell it as a mining device, you sell it as a PC with an extra feature of mining. Ive never once used a firewire device but I have owned lots of systems that have the feature. If its not something you are interested in then you dont set it up. This is the same idea as the 21inc computer. Those who view it as a mining device are missing the point.

The entire network is about 1.3 billion GH/s at the moment. 8 watts isnt going to break the bank for any homeowner, or even a large corporation I wouldnt think.

When I was a kid, I reasoned like that too: "Most people don't even pick up a penny when they drop it, they couldn't care less. If everyone in the world just sent me a penny, it "isnt going to break the bank," but I'd be a millionaire!"

Maybe this could be worded differently. Id be willing to 'donate' $0.50 a month to run the mining chip in order to help secure the network. Those who would not be can simply not configure the miner and it will sit idle.

This is uber nonsense, if the bitcoin network can't sustain itself than it has to die, no one should be subsidizing the network, and yes, you bet you're paying those extra USB ports, otherwise they wouldn't be there...

[franky1]

That's not, and should not be, an option.

A possible solution would be other variables gaining importance for mining, not just cheap electricity, for example, if blocks got bigger good Internet connections would be crucial for success when mining and the playing field could be leveled and a bigger percentage of mining would be done outside of China.

1. i agree stablizing the difficulty is bad. the higher the difficuty is the more difficult it is for a bad actor to abuse
2. it does not matter where the ASICS are run from (from the point of view of block data). asics do not store or use a full block data. they are given a small slice of data to represent the block and they has it out. its the POOLS (separate entity) that relay and hold real block data. and these pool servers DO NOT need to be at the same location as the ASIC's.

EG chinese pools can run their pool servers in america but house the ASICS in china. (they already do)

3. the long debate going on for the last year about block data is a meaningless myth.. do you think that netflix is crying, due to the false belief that only 6000 people can see their 3gb for an hour long HD show? .. no because millions of people can see 3gb in under an hour (ofcourse thats download rate)
do you think twitch is crying because only 6000 people can play an online game(upload data) while in teamspeak(upload) while livestreaming it all to twitch(upload).... no because millions of people do it.

in short internet speeds are not an issue.

[Ultrafinery]

I had this discussion with a friend not long ago and here is my thoughts... This came about when talking about 21INC and Intel being involved in blockchain tech.

There are about 2 billion PCs in the world at the moment. If each motherboard had a single ASIC chip from the AntMiner S7 (25GH/s @ 5 watts) it would equal 50billion GH/s.

Are those boxen running 24/7? Never switched off? And Toto Laptops too?
Constantly connected to the internet?
Why would anyone pay an extra, let's say $25 (because arbitrary numbers are fun)  for a computer, knowing they will "mine" $5?
Is Grandma/Visa/my hospital interested in Bitcoin mining?

Do you pay extra for your computer because it has 2 extra USB ports? or even firewire? No, its so cheap for them to add it that its just added.

Yes, I do pay extra for those USB ports. If I didn't need those USB ports, and a similar MB was available for less (without them), I would buy that instead.
If most people didn't need the extra USB ports, that MB company would soon go out of business.
That's why MB don't come with 50 USB ports -- because while the costs are negligible, the rewards are zip.
Successful businesses become successful by eliminating unnecessary costs, not adding them (as you suggest).

You dont sell it as a mining device, you sell it as a PC with an extra feature of mining.

While no one ever went broke by banking on stupidity of others, this will likely be the first time. I'm sure some people will say "Gee Honey, it's comfabulator-compatible and shnurgles SHA256, for only $25 more! OMG, let's get this!" I'm certain of this because actual, IRL people bough 21inc's $400 PiTato. The important bit here is most stayed away in droves.

The entire network is about 1.3 billion GH/s at the moment. 8 watts isnt going to break the bank for any homeowner, or even a large corporation I wouldnt think.

When I was a kid, I reasoned like that too: "Most people don't even pick up a penny when they drop it, they couldn't care less. If everyone in the world just sent me a penny, it "isnt going to break the bank," but I'd be a millionaire!"

Maybe this could be worded differently. Id be willing to 'donate' $0.50 a month to run the mining chip in order to help secure the network. Those who would not be can simply not configure the miner and it will sit idle.

Then buy yourself a USB stick. Easy peasy.
You keep comparing an extra USB port (pennies) to a miner (chip, new MB layout, likely extra VR/cooling/other support circuitry (chip != miner, other stuff there too). So apples to oranges.

If you insist on making people pay for shit they don't need, just start selling computers pre-loaded with adware, which also makes their boxen a part of your bot net. Most people wouldn't notice that either.

P.S. also see boldface, which you might've missed.

[RealBitcoin]

Even if you produce cheap mining equipment elsewhere besides China, operating costs such as electricity and manpower would still kill your profits. In all honesty the only situation that can keep mining as decentralized as possible is for the difficulty to remain stable.

Like there arent any cheap countries other than china?

I`m sure you can setup a sweathshop in any 3rd world country, but we are talking mostly about robot work here, and for  that you only need materials ,a few engineers and electricity.

For example you could setup manufacturing plants in Iceland, cheap electricity and probably less regulations.

[aerobatic]

• If the bitcoin price increases, more miners will join us due to greed and existing smaller miners will have enough profits to uppgrade, thus mining decentralizes
• Current mining equipment has reached the cutting edge of technology, the 16 nanometer chips are currently the best, and mining efficiency increase will probably slow down to a logarithmic curve now. This means that big miners wont have huge advantage anymore over little miners, by uppgrading to new stuff. Also people who buy the latest mining equipment, will suffer less deflation, and if a newer one comes out they can sell them at a better price and not lose that much money. This will keep more money in pockets of smaller miners, and they can uppgrade now more efficiently with less losses. Also miners that buy mining equipment now, can use that equipment longer before the difficulty increases thus making faster ROI and more profits so they can expand faster.
• The upcoming halving should double the price, but it can increase it even more in the longer term, making it easier for miners to mine, and decentralize it
• We must have a very good alert system and comunication system in case of a malicious hardfork happens. Especially a pre-planned one, that was due to bribery. Totalitarian takeovers of bitcoin should be fight against by the entire community!
• We must incentivize bitcoin mining again, and raise awareness to it, especially in countries where the electricity is cheap by default. People who live in cheap electricity countries must participate in bitcoin mining for the sake of bitcoin's future!
• Anyone who cares about bitcoin, should support it now, and should spread the word about it! We should stop hoping for bitcoin to succeed and start working on bitcoin's success!
• 21 Inc mining chips to be integrated in every electronic gadget, so that the masses can mine bitcoin in idle mode, decentralizing mining even more!
  https://www.cryptocoinsnews.com/breaking-21-inc-releases-new-bitcoin-mining-chip-smartphones/
  

RBC -

your grasp of the technology and economics of mining are superficial and wrong in places.

1.  Every mining chip could be integrated into electronic gadgets.  Whether its from 21 Inc or Bitfury or Antminer, there's nothing stopping anyone from doing it.  The question is whether a market exists and whether the punters want it and are willing to pay for it?  (so far it doesn't seem like the market wants it)

2.  We all care about bitcoin, all the miners, and both Core and Classic camps care about bitcoin.  Its just they have different priorities and think different solutions are whats needed to scale it (and have different incentives).   Im with Andreas Antonopolous in saying that ALL scaling solutions are needed, not just SegWit and Lightning Network for the long term but ALSO increasing the block size for the short term (...asap!)

3.  the simple fact is that mining is sensitive to electricity costs so you either have to mine in places where its cheap (like Iceland, Venezuela etc) or you have to mine in places where the electrical companies are willing to be paid a share of the bitcoin proceeds instead of fiat KwH prices (some of china/mongolia etc)

4.  16 nm is just the current process (and is the most efficient to date).  Not every mining chip is there yet.   But i don't think you grasp just how expensive it is to create a 16nm asic.  It would've cost any company that did it at least $30m.   (thats just design, NRE costs and the first few wafers)  That limits creating bitcoin mining chips to only the biggest and best financed companies.   Thats probably the biggest cause of the mining centralisation.  And its going to get worse.  9 & 10nm processes are coming next year).  And they will be even more expensive than 16nm.   The requirement to mine efficiently requires the latest custom asic chips, and that drives centralisation.  Companies that have invested significant shareholder funds to create a 16nm or next year, 10nm asics, are going to want to want to realise returns from their investment, and to do that they need to build out their own mines, and while their chip is better than everyone else's chip, they need to keep it to themselves and mine as much as they can so that they retain their cost advantage for the longest time.  They will only start selling their asic to smaller miners when their chip isn't better than anyone else's (better = lower power or lower cost) - or when the customer's pay anywhere close to the margins that they can make mining themselves.   This explains why the biggest miners in the world are the ones that make their own chips, and also explains why 21 Inc is nowhere, because  - at least historically - their chips weren't the best asics and in part explains why they try to sell them to the public.   the public, however, aren't stupid... and most of them realise they cant mine profitably nor earn back the cost of the miner they bought... thus the mining is a token effort.   the whole point of mining is to be profitable and have an incentive to do it.

5.  mining efficiency improvements are slowing down but not necessarily for the reasons you mention.  sha-256 isn't monolithic.  bit by bit, people have found ways of improving the circuit designs, both on chip and system-wide (long strings, without voltage converters).   the main reason efficiency improvements are slowing down is that previous mining chips either weren't on the latest process, or, that they weren't the most optimised designs.   but both of those things are now beginning to happen... designs are becoming more optimum (both asic design and system design)... and sha-256 algo's are getting tighter.  there's still further optimisations to be done to algorithms (for instance, asicboost.com).  the best mining chips, to date have been a combination of efficient designs, and new processes (16nm/14nm etc).  Once companies achieve the best algo's on the best processes, then they're locked into moore's law... and they're converging towards that but as yet they aren't there.  maybe next year we'll be closer tracking to moore's law, but previously we've been accelerating performance gains much faster than moore's law.

6.  agree with you that we should stop hoping and start doing.  but you haven't said anything that actually can be done.   imho we've got to stop in-fighting.  we all want bitcoin to succeed but we've got ridiculous problems that we shjouldnt be having.  We've got competing teams dissing each other.  we've got fanboys on both sides doing evil things like censorship, and down voting posts, or ridiculing people who are part of the solution.   it seems to me we're lacking a leader.  A Linus Torvalds.  its a shame Satoshi abdicated.  Maybe if he returns he can help bridge the chasm of bitcoin developers and miners.

7.  its not just people in cheap electricity places that can mine.  its also people in cold places that need heat.  that opens it up a bit more.