Bitcoin Forum

Bitcoin => Mining speculation => Topic started by: novello on December 30, 2015, 05:36:02 PM



Title: How to beat Centralisation
Post by: novello on December 30, 2015, 05:36:02 PM
Long time since last posted here, quite a lot has happened since. Basically the asic manufacturers have used revenues from sales of their mining machines to fund super sized mining farms, driving the difficulty up and up and then to add insult to injury, stopped selling their wares to consumers.

Don't want to say 'we told you so', fact is that we did but that's history, with the way things are going just now the only way for home miners to make any money is either to get 'free', (stolen, for want of a better word) electricity where essentially they are monetising the power supply or buying a new rig, running it for a few month then selling onto someone else. The rise in difficulty is going to make life harder and harder for miners and I suspect you haven't seen the worse of it yet, think 1,700+ PH by halving time next June. 

The only way individual miners are going to be able to compete is by working together, I know it's an idea that's been touted before and failed but what's needed is a community solution, one that's owned by the community and can't be abused. We tried crowdfunding and failed but that doesn't mean that it's a bad idea, especially if it was project managed by a respected figure like Gregory Maxwell, for instance. There are lots of different crowdfunding models nowadays and many of them allow for equity ownership, to fund an all new 28nm asic program - forget about 16nm for now - and use it to build a competitive super farm would cost around $20 million delivering around 300PH using 2 cents per kWh electricity. Even at 1700 PH with halving in place it could generate net profits of over $2 million a month which could be used to build more chips/systems which equity holders could buy for their own use at cost prices or add to the farm to boost their earnings.

No VC's to pay back or interfere, no manufacturers shutting off supplies just when you need more hashing power.

$20 million is a lot of money, but when you consider the number of existing and potential[/i ]miners out there that might be willing to pony up $1000 it's achievable, and many miners might want to put in a lot more, especially if they know their money isn't being used against them as it were.

Comments?


Title: Re: How to beat Centralisation
Post by: RichBC on December 30, 2015, 07:39:52 PM
So you are suggesting that the "Solution" to the current crop of ASIC / System manufactures, with Years of experience behind them, is a brand new ground up Crowd Funded ASIC / System Manufacturer, add another 300PH of Miners to the network and that somehow will save the Home Mining?

This miracle / largest pre-sale we have seen yet being performed by the time of the halving and using already outdated 28nm Technology? I think not....


Rich



Title: Re: How to beat Centralisation
Post by: novello on December 30, 2015, 08:11:15 PM
So you are suggesting that the "Solution" to the current crop of ASIC / System manufactures, with Years of experience behind them, is a brand new ground up Crowd Funded ASIC / System Manufacturer, add another 300PH of Miners to the network and that somehow will save the Home Mining?

This miracle / largest pre-sale we have seen yet being performed by the time of the halving and using already outdated 28nm Technology? I think not....


Rich



For a kickoff, 28nm isn't outdated, it's serving Bitmain very well and from the look of the current 16nm solutions they are expensive white elephants. Experience is a relative word, so far all the 16nm solutions have totally failed to meet their performance targets, as did a lot of the 28nm ones. A good custom 28nm design can easily match a cell based one in both pipelines/sq mm and J per GH/sec, and the silicon costs a lot less plus the process is a lot more mature and manufacturable.

Also, if the project happened it would have awful lot on people supporting it who had an interest in making it happen - that makes solving problems a lot easier.

Out of interest, what solution would you propose? Or is it just too much effort to try looking for one?


Title: Re: How to beat Centralisation
Post by: RichBC on December 30, 2015, 08:32:00 PM


For a kickoff, 28nm isn't outdated, it's serving Bitmain very well and from the look of the current 16nm solutions they are expensive white elephants. Experience is a relative word, so far all the 16nm solutions have totally failed to meet their performance targets, as did a lot of the 28nm ones. A good custom 28nm design can easily match a cell based one in both pipelines/sq mm and J per GH/sec, and the silicon costs a lot less plus the process is a lot more mature and manufacturable.

If your ASIC was available Today this would be a reasonable statement. However as it is just an idea, and in reality even if funding was available now would take at least 12 Months to develop, it is not an unreasonable assumption that your full custom 28nm will be looking very dated by the time it arrives.

I do not see any evidence that 14/ 16nm has failed? Yes we are only just seeing the impact but Spondoolies, Bitfury & BW plus Bitmain when they are ready will I am sure, particularly after the halving, show that this technology is essential if you want to compete.

Quote

Also, if the project happened it would have awful lot on people supporting it who had an interest in making it happen - that makes solving problems a lot easier.

Out of interest, what solution would you propose? Or is it just too much effort to try looking for one?

A solution is only required if you think there is a problem to solve? Yes it's tough as a home Miner but who says that there has to be home mining? In practice communities such as this do a great job in helping eachother, solving problems, improving efficiencies etc. Home Mining will survive but it will not compete with the Big Boys.

Even if your idea was to come to fruition it would just create another Group with a Farm owned by relatively few people who would only be interested in getting a return on their investment. The true Home miner would be left exactly where he is Today.

Rich



Title: Re: How to beat Centralisation
Post by: VirosaGITS on December 30, 2015, 08:47:11 PM


For a kickoff, 28nm isn't outdated, it's serving Bitmain very well and from the look of the current 16nm solutions they are expensive white elephants. Experience is a relative word, so far all the 16nm solutions have totally failed to meet their performance targets, as did a lot of the 28nm ones. A good custom 28nm design can easily match a cell based one in both pipelines/sq mm and J per GH/sec, and the silicon costs a lot less plus the process is a lot more mature and manufacturable.

If your ASIC was available Today this would be a reasonable statement. However as it is just an idea, and in reality even if funding was available now would take at least 12 Months to develop, it is not an unreasonable assumption that your full custom 28nm will be looking very dated by the time it arrives.

I do not see any evidence that 14/ 16nm has failed? Yes we are only just seeing the impact but Spondoolies, Bitfury & BW plus Bitmain when they are ready will I am sure, particularly after the halving, show that this technology is essential if you want to compete.

Quote

Also, if the project happened it would have awful lot on people supporting it who had an interest in making it happen - that makes solving problems a lot easier.

Out of interest, what solution would you propose? Or is it just too much effort to try looking for one?

A solution is only required if you think there is a problem to solve? Yes it's tough as a home Miner but who says that there has to be home mining? In practice communities such as this do a great job in helping eachother, solving problems, improving efficiencies etc. Home Mining will survive but it will not compete with the Big Boys.

Even if your idea was to come to fruition it would just create another Group with a Farm owned by relatively few people who would only be interested in getting a return on their investment. The true Home miner would be left exactly where he is Today.

Rich


The hashrate of the full network has almost doubled in 2 months, this is not the work of Bitmain's full custom 28nm. So its hard to say that 28nm is being very successful either.

For home mining, the only thing that could help people mine with expensive electricity rate, is heavy over/down volt capability. If you take the Outdatedish S5 chip and you allow people to undervolt to the max, you actually increase the lifetime of the software for small miner, which helps and people with cheaper electricity can overvolt instead.

For example, imagine innate ability to run the BM1384 (Antminer S5) to 0.3J/GH? It would be able to live into the "S7 Phase" no problem, at a hashrate loss. While people with cheaper electricity might run it at 0.65J/GH to get something close to 2TH/s?

For bits and reasons, i think the only way to see this happen is if "we" managed to obtain a batch of current gen chips and have Sidehack roll out full boards.

And the only way to make this happen is to crowdfund a batch.


Title: Re: How to beat Centralisation
Post by: philipma1957 on December 30, 2015, 09:17:08 PM


For a kickoff, 28nm isn't outdated, it's serving Bitmain very well and from the look of the current 16nm solutions they are expensive white elephants. Experience is a relative word, so far all the 16nm solutions have totally failed to meet their performance targets, as did a lot of the 28nm ones. A good custom 28nm design can easily match a cell based one in both pipelines/sq mm and J per GH/sec, and the silicon costs a lot less plus the process is a lot more mature and manufacturable.

If your ASIC was available Today this would be a reasonable statement. However as it is just an idea, and in reality even if funding was available now would take at least 12 Months to develop, it is not an unreasonable assumption that your full custom 28nm will be looking very dated by the time it arrives.

I do not see any evidence that 14/ 16nm has failed? Yes we are only just seeing the impact but Spondoolies, Bitfury & BW plus Bitmain when they are ready will I am sure, particularly after the halving, show that this technology is essential if you want to compete.

Quote

Also, if the project happened it would have awful lot on people supporting it who had an interest in making it happen - that makes solving problems a lot easier.

Out of interest, what solution would you propose? Or is it just too much effort to try looking for one?

A solution is only required if you think there is a problem to solve? Yes it's tough as a home Miner but who says that there has to be home mining? In practice communities such as this do a great job in helping eachother, solving problems, improving efficiencies etc. Home Mining will survive but it will not compete with the Big Boys.

Even if your idea was to come to fruition it would just create another Group with a Farm owned by relatively few people who would only be interested in getting a return on their investment. The true Home miner would be left exactly where he is Today.

Rich


The hashrate of the full network has almost doubled in 2 months, this is not the work of Bitmain's full custom 28nm. So its hard to say that 28nm is being very successful either.

For home mining, the only thing that could help people mine with expensive electricity rate, is heavy over/down volt capability. If you take the Outdatedish S5 chip and you allow people to undervolt to the max, you actually increase the lifetime of the software for small miner, which helps and people with cheaper electricity can overvolt instead.

For example, imagine innate ability to run the BM1384 (Antminer S5) to 0.3J/GH? It would be able to live into the "S7 Phase" no problem, at a hashrate loss. While people with cheaper electricity might run it at 0.65J/GH to get something close to 2TH/s?

For bits and reasons, i think the only way to see this happen is if "we" managed to obtain a batch of current gen chips and have Sidehack roll out full boards.

And the only way to make this happen is to crowdfund a batch.

I have to agree a good chip is number one priority .  Then a speed watt per gh option.

I have seen an s-7 do as well as .245 watts a gh  and as bad as .311 watts a gh.

So a bit fury with 3 watts settings

.10 ----------------  1.00 th = 100.0 watts
.15 ----------------- 1.75 th = 262.5 watts
.20 ----------------- 2.10 th = 420.0 watts

A machine like the one above would sell.



Title: Re: How to beat Centralisation
Post by: novello on December 30, 2015, 10:50:48 PM
Rather than cluttering up this this thread with quote after quote let me answer some of the inputs

1. No one knows who is adding what to the hashrate, for all anyone knows the big farms might have been building systems for the last
    6-8 months from their earnings but not deploying them in case other do too. Classic game theory behaviour, someone always goes
    against the de facto agreement. There is no evidence what the capacity added consists of, but I'll bet it isn't 16 nm.

2. A custom asic for bitcoin mining can be designed and brought to manufacture in less than 8 months with enough funding.

3. A farm initially crowdfunded will always favour it's initial investors which is why everyone who can should contribute.  Thats not to say
    that it won't sell its future systems built to non investors as their payments will help give economies of scale.

4. 28nm is far from obsolete and that will still be the case in 12 months time with a smart enough design.

5. The crowdfunded farm enterprise doesn't have to pay back VC's who expect huge returns.  It can also give it's personnel reasonable
    salaries with the need to provide useless and money wasting jobs for 'media directors' and other such vanity posts.

Good to see some discussion, though. Incidentally, this is just an idea and I'm very well aware it's unlikely to become reality.


Title: Re: How to beat Centralisation
Post by: Amph on December 31, 2015, 07:39:53 AM
it's already prevented to some extend, no matter the efficiency at some point miners will need to change territory because there is no zone with infinite electricity

a thing must be clear, if every miners right now move all their equipments in china and start work together as they were a big farm, this is still not centralization


Title: Re: How to beat Centralisation
Post by: Dalkore on December 31, 2015, 08:21:57 AM
Long time since last posted here, quite a lot has happened since. Basically the asic manufacturers have used revenues from sales of their mining machines to fund super sized mining farms, driving the difficulty up and up and then to add insult to injury, stopped selling their wares to consumers.

Don't want to say 'we told you so', fact is that we did but that's history, with the way things are going just now the only way for home miners to make any money is either to get 'free', (stolen, for want of a better word) electricity where essentially they are monetising the power supply or buying a new rig, running it for a few month then selling onto someone else. The rise in difficulty is going to make life harder and harder for miners and I suspect you haven't seen the worse of it yet, think 1,700+ PH by halving time next June.  

The only way individual miners are going to be able to compete is by working together, I know it's an idea that's been touted before and failed but what's needed is a community solution, one that's owned by the community and can't be abused. We tried crowdfunding and failed but that doesn't mean that it's a bad idea, especially if it was project managed by a respected figure like Gregory Maxwell, for instance. There are lots of different crowdfunding models nowadays and many of them allow for equity ownership, to fund an all new 28nm asic program - forget about 16nm for now - and use it to build a competitive super farm would cost around $20 million delivering around 300PH using 2 cents per kWh electricity. Even at 1700 PH with halving in place it could generate net profits of over $2 million a month which could be used to build more chips/systems which equity holders could buy for their own use at cost prices or add to the farm to boost their earnings.

No VC's to pay back or interfere, no manufacturers shutting off supplies just when you need more hashing power.

$20 million is a lot of money, but when you consider the number of existing and potential[/i ]miners out there that might be willing to pony up $1000 it's achievable, and many miners might want to put in a lot more, especially if they know their money isn't being used against them as it were.

Comments?

It saddens me to make this statement, of all the wonderfulness that Satoshi gave the world with Bitcoin he has one fatal error in the concept that I am not sure we can fix in Bitcoin 1.0.   This may offend people or they may just disagree and I can accept that but here it is:

You can not secure a network of trust and confidence on human greed.  Eventually it will consume itself to the point where only the most aggressive, ruthless and cunning players get to reap most if not all the benefits.  Satoshi had epic insight when he/she/group/x made Bitcoin but with the great understanding of economics, there is a glaring lack of understanding of human nature on this one point.  Bitcoin is not robust in this area and in the end it the mining/processing will only be done by a few huge conglomerate interests and that is it.  


You do not know how much this pains me to say.  The best we can do it try and put capital into people who are doing what they can to make this accessible as long as they can but in the end, mining is not and will not be for the people in a manner that they matter at all, full stop.

I(Dalkore), personally am on record a long time ago warning of this before ASICs hit and I was decried as either stupid or a hater.  Well now I am being proven correct and it gives me zero pleasure in being right.  Human interest and greed are some of the most powerful stuff around, it literally can move mountains if directed.  In this case it was directed at the mining of a p2p decentralized digital currency and look what we have created, I literally have a data-center that only exist because of this singular idea that Satoshi gave to the world.  

I don't know what a solution could be and there might not even be one expect let it runs its course and maybe that will give us insight on how to fix it.  

On the bright side, the idea of a p2p decentralized digital currency with a publicly shared ledger is now planted firmly in the human mind and THAT, will never go away and even more epic ideas will spawn from this, you can be assured of that.  

If you want to talk about this more in-depth, you know how you can reach me.    Thank you.

-Dalkore


Title: Re: How to beat Centralisation
Post by: dime2spend on December 31, 2015, 10:53:50 AM
Quote
You can not secure a network of trust and confidence on human greed.  Eventually it will consume itself to the point where only the most aggressive, ruthless and cunning players get to reap most if not all the benefits.  Satoshi had epic insight when he/she/group/x made Bitcoin but with the great understanding of economics, there is a glaring lack of understanding of human nature on this one point.  Bitcoin is not robust in this area and in the end it the mining/processing will only be done by a few huge conglomerate interests and that is it. 

I fully agree with that. In fact that statement is true for many Bonus Systems. Giving a bonus for a certain behaviour is usually an invitation for the greedy ones to seek ways to maximize their profits. Malus Systems usually work better. You get punished for unwanted behaviour so you either follow the rules or you have to pay a fee. Regarding Bitcoins that could mean you have to pay a fee for a transaction if you don't run a node (with some minimum mining hashing capabilities). Something like that could still save Bitcoin in the long run.


Title: Re: How to beat Centralisation
Post by: Korbman on December 31, 2015, 02:35:32 PM
You can not secure a network of trust and confidence on human greed.  Eventually it will consume itself to the point where only the most aggressive, ruthless and cunning players get to reap most if not all the benefits.  [...]

100% this. So long as mining revenue (from any source, past or present...CPUs, GPUs, FPGAs, ASICS, etc) exceeds the cost to generate that revenue, mining power will continue to be centralized.

And why wouldn't it? If you have a piece of hardware that earns you $1, but you find out you can earn $2 if you move it to an area with slightly cheaper energy costs, wouldn't you? Continue that trend on scale - develop more efficient hardware, move that hardware to cheap power sources, find ways to reduce any and all expenses, and suddenly you'll find yourself operating a monstrous, profitable mine that no amateur can compete against.

The only way to stop that trend is if there's some sort of financial disincentive to consolidation.


Title: Re: How to beat Centralisation
Post by: baazju on December 31, 2015, 04:18:04 PM
Back when I started mining (3-4 years ago) people were saying these same things. Home mining is dead don't even bother. I quit for a year just as the first ASICs were coming out because of all of this speculation. And of course home mining is still not dead. I hate centralization and wish we could go back to the days of CPU/GPU home miners, but even now there is still hope for home miners even if it for just fun and not profit.

Personally I'll mine my current hardware until it dies whether I make money or not. I have enough BTC now that as long as the value keeps growing I'll be happy.


Title: Re: How to beat Centralisation
Post by: VirosaGITS on December 31, 2015, 07:51:40 PM
Back when I started mining (3-4 years ago) people were saying these same things. Home mining is dead don't even bother. I quit for a year just as the first ASICs were coming out because of all of this speculation. And of course home mining is still not dead. I hate centralization and wish we could go back to the days of CPU/GPU home miners, but even now there is still hope for home miners even if it for just fun and not profit.

Personally I'll mine my current hardware until it dies whether I make money or not. I have enough BTC now that as long as the value keeps growing I'll be happy.

"You" should never listen to other people when they tell you "Yes you can do this." or "No you can't do this anymore." BTC mining is ruled by your electricity price and then your hardware acquisitions costs.

This is very simple math and listening to people that tell you yes, because it is for them or no, because it is not for them is not going to help "you".


Title: Re: How to beat Centralisation
Post by: baazju on December 31, 2015, 08:02:48 PM
Back when I started mining (3-4 years ago) people were saying these same things. Home mining is dead don't even bother. I quit for a year just as the first ASICs were coming out because of all of this speculation. And of course home mining is still not dead. I hate centralization and wish we could go back to the days of CPU/GPU home miners, but even now there is still hope for home miners even if it for just fun and not profit.

Personally I'll mine my current hardware until it dies whether I make money or not. I have enough BTC now that as long as the value keeps growing I'll be happy.

"You" should never listen to other people when they tell you "Yes you can do this." or "No you can't do this anymore." BTC mining is ruled by your electricity price and then your hardware acquisitions costs.

This is very simple math and listening to people that tell you yes, because it is for them or no, because it is not for them is not going to help "you".

I should say that the speculation was just one of many reasons of why I stopped, there were other non-bitcoin related reasons as well. But when I stopped no one really knew the math to figure out what impact the ASICs would have. Or at least not in any of the communities I was hanging around in. Everyone seemed to think it would make home mining worthless and there was "math" to show it.

The math isn't really easy though. You could try and see what would be worth doing right now. But who knows what the price of BTC will be in a month or even a week.


Title: Re: How to beat Centralisation
Post by: VirosaGITS on December 31, 2015, 10:54:13 PM
Back when I started mining (3-4 years ago) people were saying these same things. Home mining is dead don't even bother. I quit for a year just as the first ASICs were coming out because of all of this speculation. And of course home mining is still not dead. I hate centralization and wish we could go back to the days of CPU/GPU home miners, but even now there is still hope for home miners even if it for just fun and not profit.

Personally I'll mine my current hardware until it dies whether I make money or not. I have enough BTC now that as long as the value keeps growing I'll be happy.

"You" should never listen to other people when they tell you "Yes you can do this." or "No you can't do this anymore." BTC mining is ruled by your electricity price and then your hardware acquisitions costs.

This is very simple math and listening to people that tell you yes, because it is for them or no, because it is not for them is not going to help "you".

I should say that the speculation was just one of many reasons of why I stopped, there were other non-bitcoin related reasons as well. But when I stopped no one really knew the math to figure out what impact the ASICs would have. Or at least not in any of the communities I was hanging around in. Everyone seemed to think it would make home mining worthless and there was "math" to show it.

The math isn't really easy though. You could try and see what would be worth doing right now. But who knows what the price of BTC will be in a month or even a week.

Well i said "You" to not mean you specifically, so that it can apply to anyone, but its true that it is much easier to do now than before because things are much more predictable now.

Its not really hard to run the math at -30% BTC value or +30% BTC value. I generally run my scenario by "Worse case scenario, can i break even?" If yes, odds are i'm going to break even + any extra is profits.

Like i setup my mining to profit at BTC price at 210$ BTC value, when we were at 240$. Now we've doubled, but the difficulty pretty much doubled as well by now, though up until now, it was gravy, now i'm back to about same profitability risk as foreseeing 210$ worse case scenario then, if BTC drops.

Of course the true worse case scenario is the world exploding, BTC dropping to 0$, but i don't think it can happen unless something globally dramatic happen, so i feel "safe enough".


Title: Re: How to beat Centralisation
Post by: Bepesand on January 02, 2016, 08:39:50 PM
Back when I started mining (3-4 years ago) people were saying these same things. Home mining is dead don't even bother. I quit for a year just as the first ASICs were coming out because of all of this speculation. And of course home mining is still not dead. I hate centralization and wish we could go back to the days of CPU/GPU home miners, but even now there is still hope for home miners even if it for just fun and not profit.

Personally I'll mine my current hardware until it dies whether I make money or not. I have enough BTC now that as long as the value keeps growing I'll be happy.

Home mining is dead for the people in high electricity cost countries such as UK and Germany.