Author

Topic: ASICMINER: Entering the Future of ASIC Mining by Inventing It - page 695. (Read 3917635 times)

hero member
Activity: 784
Merit: 501
I don't fear this will be a big issue by then, when Bitcoin becomes way more popular energy efficiency will become a more important factor and chips will have to become more energy efficient if the miners want to make a profit.

THats incorrect. It doesnt matter how energy efficient you make them, once these hardware prices have come down to something near marginal cost, electricity cost will be the overriding factor for miners, and the only real limit on network growth. More efficient chips would just result lower mining cost per TH, which will  results in a proportionally higher network speed, and thus rendering the higher efficiency pointless. Have a look here:
https://bitcointalksearch.org/topic/finally-a-correct-endgame-difficulty-calculator-295270

For a given BTC price and electricity cost, you can pretty much calculate how many megawatt bitcoin will consume. The only way to prevent that would be if hardware became more expensive somehow, so that the hardware investment would be a bigger brake. Not very likely with asics.

If enough people take their miners offline because they don't make a profit difficulty will drop and then it will be (briefly) profitable to mine again. It then helps if you have the best energy efficiency. I guess when the limit is reached we will see cycles between high difficulty and low difficulty ad infinitum, until quantum computers come online and the arms race starts again.
legendary
Activity: 980
Merit: 1040
I don't fear this will be a big issue by then, when Bitcoin becomes way more popular energy efficiency will become a more important factor and chips will have to become more energy efficient if the miners want to make a profit.

THats incorrect. It doesnt matter how energy efficient you make them, once these hardware prices have come down to something near marginal cost, electricity cost will be the overriding factor for miners, and the only real limit on network growth. More efficient chips would just result in almost proportionally lower mining cost per TH, which will  results in a proportionally higher network speed, and thus rendering the higher efficiency pointless. Have a look here:
https://bitcointalksearch.org/topic/finally-a-correct-endgame-difficulty-calculator-295270

For a given BTC price and electricity cost, you can pretty much calculate how many megawatt bitcoin will consume. The only way to prevent that would be if hardware became more expensive somehow, so that the hardware investment would be a bigger brake. Not very likely with asics.
hero member
Activity: 784
Merit: 501
The entire bitcoin network doesnt really represent a whole lot yet does it? Not compared to PP or banks, heck not even compared to Bloomberg.
I wouldnt go as far as calling it a environmental threat, but its certainly not something to take pride in and if bitcoin ever becomes a mainstream currency, it would get fairly ugly.

I don't fear this will be a big issue by then, when Bitcoin becomes way more popular energy efficiency will become a more important factor and chips will have to become more energy efficient if the miners want to make a profit.
legendary
Activity: 980
Merit: 1040
Banks have datacenters that use up more energy than the Bitcoin network.

Banks offer services and volume that is just not comparable to the blockchain.
Paypal is a much more direct comparison, even though they do a lot more than just processing transactions, and I just showed its using a comparable amount of energy. Except PP handles $100 billion worth of payments each year for 130 million users and does tons of other things too. That makes it very much more energy efficient.

Quote
There was an article on Bloomberg that called Bitcoin a realworld threat to the environment, I read a rebuttal on one of the Bitcoin news sites (can't remember which) in which the author had calculated that the Bloomberg HQ building in New York used more energy than the entire Bitcoin network. It is really a red herring.

The entire bitcoin network doesnt really represent a whole lot yet does it? Not compared to PP or banks, heck not even compared to Bloomberg.
I wouldnt go as far as calling it a environmental threat, but its certainly not something to take pride in and if bitcoin ever becomes a mainstream currency, it would get fairly ugly.



hero member
Activity: 784
Merit: 501
The argument that Bitcoin uses alot of energy is just a cheap shot that is easily refuted with comparison to the current systems. You don't even to go deep and comprehensive like the above just the ATM machines woeldwide and you are done.

Believe it or not some people bring that up when we are talking bitcoin and they feel like undermining the system, pun intended  Cheesy

Are you saying bitcoin ATMs are more energy efficient, or that bitcoin somehow eliminates the need for POS transactions?

Banks have datacenters that use up more energy than the Bitcoin network. There was an article on Bloomberg that called Bitcoin a realworld threat to the environment, I read a rebuttal on one of the Bitcoin news sites (can't remember which) in which the author had calculated that the Bloomberg HQ building in New York used more energy than the entire Bitcoin network. It is really a red herring.
legendary
Activity: 980
Merit: 1040
The argument that Bitcoin uses alot of energy is just a cheap shot that is easily refuted with comparison to the current systems. You don't even to go deep and comprehensive like the above just the ATM machines woeldwide and you are done.

Believe it or not some people bring that up when we are talking bitcoin and they feel like undermining the system, pun intended  Cheesy

Are you saying bitcoin ATMs are more energy efficient, or that bitcoin somehow eliminates the need for POS transactions?
hero member
Activity: 924
Merit: 1003
Unlimited Free Crypto
Now where did I see a comparison that said that one major bank central office used probably more electricity than all the bitcoin mining in the world ...

That can't be true, I would be surprised
OK lets try a wild estimate Smiley

I use less than 1KW to do well over 100GH/s
So lets go with 1KW for 100GH/s
That's a way over-estimate for some of the high hashing devices and a way under-estimate for any of the old devices anyone is silly enough to keep mining with

The bitcoin network is ~1PH/s
So that's like 10,000KW

That seem too high for a big bank's main data centre head office?
If it is then say how about 10 of them? Certainly not too high for 10 of them.

All guesses, but certainly makes that argument above seem far from certain.

That's 10MW, iirc any one of the top 50 supercomputers uses at least 15MW. Just the ATM machines worldwide consumed several times the power of the Bitcoin network too.

The argument that Bitcoin uses alot of energy is just a cheap shot that is easily refuted with comparison to the current systems. You don't even to go deep and comprehensive like the above just the ATM machines woeldwide and you are done.

Believe it or not some people bring that up when we are talking bitcoin and they feel like undermining the system, pun intended  Cheesy
legendary
Activity: 1190
Merit: 1001
I went through old friedcat posts and realized that last three updates in this thread were posted month apart. And now it's around one month since last one. I guess we should expect update any day now. (last posts were june 21st, july 23rd and august 22nd)

I hope he also come with some news about future plan and gen2 chips.
sr. member
Activity: 406
Merit: 250
I went through old friedcat posts and realized that last three updates in this thread were posted month apart. And now it's around one month since last one. I guess we should expect update any day now. (last posts were june 21st, july 23rd and august 22nd)
legendary
Activity: 980
Merit: 1040
OK lets try a wild estimate Smiley

I use less than 1KW to do well over 100GH/s
So lets go with 1KW for 100GH/s
That's a way over-estimate for some of the high hashing devices and a way under-estimate for any of the old devices anyone is silly enough to keep mining with

The bitcoin network is ~1PH/s
So that's like 10,000KW

That seem too high for a big bank's main data centre head office?
If it is then say how about 10 of them? Certainly not too high for 10 of them.

All guesses, but certainly makes that argument above seem far from certain.

Well lets compare, not to a bank, but to paypal/ebay.

The first phase of the Topaz project is a 240,000 square foot building housing three 20,000 square foot data center halls – one for eBay Marketplace, one for PayPal.com, and a third hall for expansion space. The master plan for the site calls for four phases, which will allow eBay to consolidate leased data center space currently spread across three states. The facility has 7.2 megawatts of capacity in phase 1, with a 30 megawatt substation on site.
http://www.datacenterknowledge.com/archives/2010/05/23/ebay-unveils-new-flagship-data-center/

So bitcoin is currently using a roughly comparable amount of electricity as ebay+paypal datacenter.
If you are going to spread that over  transaction volumes, honestly bitcoin isnt going to look very green.
full member
Activity: 160
Merit: 100
The power used to sustain either currency will be negligible compared to the effect on power consumption due to the economic effects on investment.

The question is really which currency system promotes investment with a meaningful real rate of return.

All of which is off-topic.
legendary
Activity: 4634
Merit: 1851
Linux since 1997 RedHat 4
Now where did I see a comparison that said that one major bank central office used probably more electricity than all the bitcoin mining in the world ...

That can't be true, I would be surprised
OK lets try a wild estimate Smiley

I use less than 1KW to do well over 100GH/s
So lets go with 1KW for 100GH/s
That's a way over-estimate for some of the high hashing devices and a way under-estimate for any of the old devices anyone is silly enough to keep mining with

The bitcoin network is ~1PH/s
So that's like 10,000KW

That seem too high for a big bank's main data centre head office?
If it is then say how about 10 of them? Certainly not too high for 10 of them.

All guesses, but certainly makes that argument above seem far from certain.
hero member
Activity: 924
Merit: 1003
Unlimited Free Crypto
Banks love to print papers and send letters, that's a lot of energy consumption. Wink

We are so arrogant to make our own currency that is created at a limited amount within a range of time and in the process use electricity.

We should let them print infinite money with infinity paper and electricity cost. That is better for the economy and the environment  Cool.

We can't be thinking better it is not like we have brains to think with, right Satoshi?
sr. member
Activity: 302
Merit: 250
Now where did I see a comparison that said that one major bank central office used probably more electricity than all the bitcoin mining in the world ...

That can't be true, I would be surprised
legendary
Activity: 4634
Merit: 1851
Linux since 1997 RedHat 4
Now where did I see a comparison that said that one major bank central office used probably more electricity than all the bitcoin mining in the world ...
hero member
Activity: 630
Merit: 500
Bitgoblin
I shudder to think of the electricity wasted otherwise.
I'm quite concerned about scaling too, but apparently most people aren't.


Can you copy paste what you wrote there here? It seems to be a private forum.


Sure!

I can copy what I wrote, but I can't copy other people's answers, though.

Quote
The most compelling argument in defence of the resources spent for the proof-of-work is that the current legacy systems (banks + credit cards + paypals + etc.) are already eating up tons of resources, and likely mining bitcoins is going to eat much less than that.

While this appears to be fine, I've just noticed it doesn't really work: the legacy systems scale in a more or less proportional way to the transaction volume, i.e. if transaction volume doesn't increase, they do not need to scale up.

Bitcoin's proof-of-work, instead, will apparently go up indefinitely, because either miners will move to countries where electricity and sysadmins are cheaper, or they will find ways to produce more powerful chips. Problem here isn't only the consumed electricity, but the wasted hardware: old chips will be useless and will become just pollution.

Honestly I can't see how can this not go horribly wrong. We'd better think very hard about this now that we are still in early stage, rather than waiting for it to become a huge problem later.
hero member
Activity: 728
Merit: 500
Yes this is variance. You can verifiy it after 3 days...

The whole network suffers from bad luck  sometimes. We could have 1400.093 Thash/s right now or 886.093 Thash/s... Its just variance



You mean you can verify its not variance only after 3 days?

That's just a figure plucked from the ether. All you can say is the longer the higher hashing rate appears to continue, the less likely it is due to variance. You can't put a specific time limit on it. This is luck we are talking about, as we keep saying.
No.

There is a reason why the difficulty adjustment is (14 days) ... 2016 blocks ...

Sorry, lost me, the guy said "you can verify after 3 days" and you say "There is a reason why the difficulty adjustment is (14 days) ... 2016 blocks ..." - how are the two connected?

The reason the difficulty adjustment is every 14 days (approximately) is that you need a period of that length (or longer) to reduce potential variance to reasonable levels. Adjusting the difficulty every day would cause large swings in both directions.

So if you already need 14 days to get a reasonable average of the network hashrate, you'll need more than that to properly reduce variance for a part of the network (in this case, ASICMiner), which brings us back to the main point that any hashrate estimate over a period of 3 days or less is completely meaningless due to the random nature of block-solving. Even longer periods, like 1 week or 2 weeks are susceptible enough to variance that it can be hard to draw any conclusions from those data.
hero member
Activity: 518
Merit: 500
Yes this is variance. You can verifiy it after 3 days...

The whole network suffers from bad luck  sometimes. We could have 1400.093 Thash/s right now or 886.093 Thash/s... Its just variance



You mean you can verify its not variance only after 3 days?

That's just a figure plucked from the ether. All you can say is the longer the higher hashing rate appears to continue, the less likely it is due to variance. You can't put a specific time limit on it. This is luck we are talking about, as we keep saying.
No.

There is a reason why the difficulty adjustment is (14 days) ... 2016 blocks ...

Sorry, lost me, the guy said "you can verify after 3 days" and you say "There is a reason why the difficulty adjustment is (14 days) ... 2016 blocks ..." - how are the two connected?
legendary
Activity: 4634
Merit: 1851
Linux since 1997 RedHat 4
Yes this is variance. You can verifiy it after 3 days...

The whole network suffers from bad luck  sometimes. We could have 1400.093 Thash/s right now or 886.093 Thash/s... Its just variance



You mean you can verify its not variance only after 3 days?

That's just a figure plucked from the ether. All you can say is the longer the higher hashing rate appears to continue, the less likely it is due to variance. You can't put a specific time limit on it. This is luck we are talking about, as we keep saying.
No.

There is a reason why the difficulty adjustment is (14 days) ... 2016 blocks ...
hero member
Activity: 518
Merit: 500
Yes this is variance. You can verifiy it after 3 days...

The whole network suffers from bad luck  sometimes. We could have 1400.093 Thash/s right now or 886.093 Thash/s... Its just variance



You mean you can verify its not variance only after 3 days?

That's just a figure plucked from the ether. All you can say is the longer the higher hashing rate appears to continue, the less likely it is due to variance. You can't put a specific time limit on it. This is luck we are talking about, as we keep saying.
Jump to: