ASICs are the reason the price is dropping, and it will not stop.

[Coinseeker]

Very interesting thread.  Makes me wonder why people, other than those who have sunk thousands into mining rigs, hold tight to the whole mining process at all.  Premine -> Consensus -> Problem solved.

Are you suggesting there's a better method than mining?  It's like democracy, it's the best of all available options, even if it's shit.

I'd argue the only better way is to use an algorithm that would somehow be inherently low barrier to entry and harder to centralize, maybe like scrypt, but preferably CPU only or something

There's always a better way.  Eliminate mining altogether.  The only purpose is to prevent double spending and that can be achieved via consensus as Ripple is doing.  The obvious downside is how to fairly distribute the currency.  A challenge yes, but one that is easier to solve than the compounding mining problems, that will ultimately destroy the currency anyway.  At least, that's how I'm reading your posts.  I could be way off on my understanding though.

[ecliptic]

ASICs are made & bought with one goal in mind: Profit.  If those rigs are profitable in dollars, it's stupid for the manufacturers to sell them.  If you have something guaranteed to make you 2 dollars by tomorrow, you don't sell it for a dollar today.  Not unless you're strung & need a fix.  So, unless ASIC manufacturers are junkies or fools, they're selling their rigs because:
A.  They know the rigs won't recoup their cost in BTCBTCBTC with the number of them on the market & accompanying growth in difficulty.
B.  They know the rigs will recoup their cost in BTCBTCBTC, but by that time the value of BTCBTCBTC will fall, so the $$$ cost won't be recouped.

Not sure what the other alternatives are, considering no one seems to need a fix.

You're forgetting that selling hardware is legally, financially, and technically much, much dimplier than operating a miner which could conceivably by an international money transmitter, have to go through AML laws, all that bullshit.

I thought Asicminer was mining online already -- so the expertise & legal liability's there already, scaling doesn't seem to be much of a problem (if you're printing $$$, you're f8cked about the same if you're busted for 100,000 or 100 mil.)
 

Selling chips is a significantly safer, simplier and better way to make money.  And FAR less risk, even just economically speaking.  Avalon makes >1000% profit on every chip they sell

True, once the r&D & tooling up's done, it's all profit.  I think you've missed my point.  Maybe not.

Also, they'd have to compete with ASICminer, who has 1000 TH/sec coming online by the end of 2013.  Not easy to do.

Thought they were already up & mining?

they are, they operate out of china i believe, and are trying to set up off-shore operations because they're worried about the authorities

indeed, governments are a risk to ASICminer.  But bribes can work well in this regard

[Walsoraj]

IMHO these ASIC-folk have ultra-incentive to purchase loads of another pumpable coin w/ same hardware (thinking PPC, as ltc is asic resistant atm) before a 51-atk on BTC... Lots of USD to be made doing that...

Has anyone thought of this? What are the protections?

[ecliptic]

Very interesting thread.  Makes me wonder why people, other than those who have sunk thousands into mining rigs, hold tight to the whole mining process at all.  Premine -> Consensus -> Problem solved.

Are you suggesting there's a better method than mining?  It's like democracy, it's the best of all available options, even if it's shit.

I'd argue the only better way is to use an algorithm that would somehow be inherently low barrier to entry and harder to centralize, maybe like scrypt, but preferably CPU only or something

There's always a better way.  Eliminate mining altogether.  The only purpose is to prevent double spending and that can be achieved via consensus as Ripple is doing.  The obvious downside is how to fairly distribute the currency.  A challenge yes, but one that is easier to solve than the compounding mining problems, that will ultimately destroy the currency anyway.  At least, that's how I'm reading your posts.  I could be way off on my understanding though.

How do you do it without centralization?

[mgio]

Remember when people said that rise in hashrate caused a rise in price?

Now everyone is saying the opposite.

The reality is the hashrate has no effect on price whatsoever!!

You assume all miners have the same intention of coins.

GPU miners held coins.

ASIC miners sell coins.

The market is so small that the 3600 coins per day mined has a massive effect on it.

How can you make this generalization?

Why would ASIC miners sell their coins more than GPU miners?

Remember that BFL customers bought their 60GH/s miners for just $1299 (including shipping!). That's the same price as a pretty decent GPU mining setup. The Jalapenos were roughly 1/10th that price.

Avalon batch 1 paid about that much too. Avalon batch 2 was just a little more expensive depending on when you bought.

Why would these miners be more despearte to sell their coins?

I mine with ASICs (my first Avalon arrives on Monday) and I have no intention of selling off my coins, at least not for while.

[Coinseeker]

Very interesting thread.  Makes me wonder why people, other than those who have sunk thousands into mining rigs, hold tight to the whole mining process at all.  Premine -> Consensus -> Problem solved.

Are you suggesting there's a better method than mining?  It's like democracy, it's the best of all available options, even if it's shit.

I'd argue the only better way is to use an algorithm that would somehow be inherently low barrier to entry and harder to centralize, maybe like scrypt, but preferably CPU only or something

There's always a better way.  Eliminate mining altogether.  The only purpose is to prevent double spending and that can be achieved via consensus as Ripple is doing.  The obvious downside is how to fairly distribute the currency.  A challenge yes, but one that is easier to solve than the compounding mining problems, that will ultimately destroy the currency anyway.  At least, that's how I'm reading your posts.  I could be way off on my understanding though.

How do you do it without centralization?

1.  Figure out how to fairly distribute all the limited currency
2.  Make it open source

I'm not saying 1. is easy, but with all these smart people there's certainly a way...one that would be far easier to solve than all of this mining nonsense.

[ecliptic]

Very interesting thread.  Makes me wonder why people, other than those who have sunk thousands into mining rigs, hold tight to the whole mining process at all.  Premine -> Consensus -> Problem solved.

Are you suggesting there's a better method than mining?  It's like democracy, it's the best of all available options, even if it's shit.

I'd argue the only better way is to use an algorithm that would somehow be inherently low barrier to entry and harder to centralize, maybe like scrypt, but preferably CPU only or something

There's always a better way.  Eliminate mining altogether.  The only purpose is to prevent double spending and that can be achieved via consensus as Ripple is doing.  The obvious downside is how to fairly distribute the currency.  A challenge yes, but one that is easier to solve than the compounding mining problems, that will ultimately destroy the currency anyway.  At least, that's how I'm reading your posts.  I could be way off on my understanding though.

How do you do it without centralization?

1.  Figure out how to fairly distribute all the limited currency
2.  Make it open source

I'm not saying 1. is easy, but with all these smart people there's certainly a way...one that would be far easier to solve than all of this mining nonsense.

How do you do the "consensus" that ripple does without centralizing it i mean.

Bitcoin/*coin does the consensus by mining power.  But without block or transaction fees, there's no incentive to run the mining network. 

[tutkarz]

Very interesting thread.  Makes me wonder why people, other than those who have sunk thousands into mining rigs, hold tight to the whole mining process at all.  Premine -> Consensus -> Problem solved.

Are you suggesting there's a better method than mining?  It's like democracy, it's the best of all available options, even if it's shit.

I'd argue the only better way is to use an algorithm that would somehow be inherently low barrier to entry and harder to centralize, maybe like scrypt, but preferably CPU only or something

There's always a better way.  Eliminate mining altogether.  The only purpose is to prevent double spending and that can be achieved via consensus as Ripple is doing.  The obvious downside is how to fairly distribute the currency.  A challenge yes, but one that is easier to solve than the compounding mining problems, that will ultimately destroy the currency anyway.  At least, that's how I'm reading your posts.  I could be way off on my understanding though.

How do you do it without centralization?

1.  Figure out how to fairly distribute all the limited currency
2.  Make it open source

I'm not saying 1. is easy, but with all these smart people there's certainly a way...one that would be far easier to solve than all of this mining nonsense.

1. is impossible so please stop this ripple propaganda here. nobody wants it.

[Impaler]

I think were just seeing the deflation of the bubble as hype has died down the amount of dollars offered on exchanges to buy coins has declined precipitously.  This graph is the single most important graph for understanding BTC.

http://www.bitcoinx.com/charts1/depth_mtgox.png

The value of BTC is almost always equal to the dollar depth divided by the coin depth, as coin depth increases or dollar depth declines then the exchange rate will fall.  Dollar depth peeked in mid May and has been on a steady decline punctuated by some sharp drops and then a temporary bulge during the Gox withdraw black-out.  Coin depth was in decline right up to the day of the peak and has been rising steadily since late May.

In the end the value of BTC is determined by the daily flow of dollars willing to invest permanently in BTC.  At the start of the year before the latest bubble that figure was around $100,000 and maintained a price of ~$15.  The sustainable price is directly proportional to the sustained investment rate divided by coins offered, but it is very likely that only a tiny fractions of new coins are being offered for sale and virtually no old coins.  Their could be some effect if ASIC miners are more or less inclined to sell their coins then GPU miners but if anything I suspect GPU miners had to sell more due to their high marginal electricity costs and the primary driver of current exchange rate declines are on the demand side.

This is a question that is amenable to study, just identify coins mined by ASIC miner and compare how many of them have passed through an exchange vs the remaining miners.

[ecliptic]

So thinking about it the real issue is that the middle man markup on ASICs from TSMC, as well as the massive economies of scale on PCB assemblies, means it's exceptionally difficult if not impossible to come anywhere near AM on hardware.

That's not even to get into their electricity, so they currently have ~110nm asics which are less efficient than the 65/28nm that should be hitting the market, but i bet they pay ~0.03c/kWh which pretty much evens the playing field.

If they go to 65nm or lower, it will probably be to make their $ / TH/sec cheaper.

[mmeijeri]

The obvious downside is how to fairly distribute the currency.

It isn't obvious that mining is a fair way to do it, or that fairness is even a problem. You can't continue to sell something for more than it's worth.

[w00t]

Thanks for the original post - I was wondering why is the price going down for some time...

As I read through the whole thread - would not the solution to this problem be if we changed the alghorithm behind Bitcoin so those ASIC can't mine it and everyone had to downgrade to the CPU or at best GPU level?

I thought that this is built-in procedure if anything goes wrong (ie. 51% attack, some vulnerability found etc..) and OP says ASIC is waaaaay over the 51% atm so it might be a good idea to put ASICs out of the game.

[mmeijeri]

As I read through the whole thread - would not the solution to this problem be if we changed the alghorithm behind Bitcoin so those ASIC can't mine it and everyone had to downgrade to the CPU or at best GPU level?

See this thread:

defending ahead the p2p nature of bitcoin - blending hashcash & scrypt

There's no need to push out the ASICs, just to make sure Bitcoin stays distributed, instead of serving the interests of a small group. The existence of ASICs isn't inherently bad, they can prevent other forms of misbehaviour, such as a voting-based Bitcoin that imposes taxes.

[bitleif]

I'm so sick of these "price is dropping because of X" bullshit threads. Price is dropping because it had no business being $150+ in the first place. It's way too early for that kind of valuation compared to actual current use as currency. It was a BUBBLE, and bubbles always pop.

[Coinseeker]

How do you do the "consensus" that ripple does without centralizing it i mean.

It's my understanding that centralization only occurs if the network is closed source.  The only trick - and granted its a big one - is how to fairly distribute all the currency.  

Bitcoin/*coin does the consensus by mining power.  But without block or transaction fees, there's no incentive to run the mining network.  

Exactly and as soon as it's no longer profitable to mine, people will stop mining and the network will collapse. Remove the need for mining and you remove all the negatives that go along with that.  Consensus in Ripple's case (which just happens to be the only one to reference to) is done through the nodes on the network being in agreement of the ledger, thus mining is no longer necessary.I'm not saying that the answers are all figured out.  I'm simply saying there are always better ways to achieve a goal and conceding to mining as if there is no other way, defies ones belief in human ingenuity.  

[w00t]

As I read through the whole thread - would not the solution to this problem be if we changed the alghorithm behind Bitcoin so those ASIC can't mine it and everyone had to downgrade to the CPU or at best GPU level?

See this thread:

defending ahead the p2p nature of bitcoin - blending hashcash & scrypt

There's no need to push out the ASICs, just to make sure Bitcoin stays distributed, instead of serving the interests of a small group. The existence of ASICs isn't inherently bad, they can prevent other forms of misbehaviour, such as a voting-based Bitcoin that imposes taxes.

Thank you for posting the link to that discussion - I read through the topic and I understood some of it (it's far too crypto-technical for me even thought I'd say I'm a power user).

If I understood it correctly there is intention to do a split so two alghorithms are used at once at rate of 50:50? It might not be that bad idea, right?

[w00t]

I'm so sick of these "price is dropping because of X" bullshit threads. Price is dropping because it had no business being $150+ in the first place. It's way too early for that kind of valuation compared to actual current use as currency. It was a BUBBLE, and bubbles always pop.

If you are sick of it then why are you here reading it?

What is in your opinion fair market price then?

Edit: Btw. unless I hear from you a reasonable reply you go to my special list.

[mmeijeri]

If I understood it correctly there is intention to do a split so two alghorithms are used at once at rate of 50:50? It might not be that bad idea, right?

It sounds like a good idea to me, though it is too early to tell whether it will be necessary. Right now it's probably not a viable option, because most Bitcoin users would prefer not to have a fork and because the fact that it isn't self-evidently necessary would likely cause a fork.

[ecliptic]

Thanks for the original post - I was wondering why is the price going down for some time...

As I read through the whole thread - would not the solution to this problem be if we changed the alghorithm behind Bitcoin so those ASIC can't mine it and everyone had to downgrade to the CPU or at best GPU level?

I thought that this is built-in procedure if anything goes wrong (ie. 51% attack, some vulnerability found etc..) and OP says ASIC is waaaaay over the 51% atm so it might be a good idea to put ASICs out of the game.

Changing the protocol isn't an option, that just kills the coin entirely. 

[crumbs]

I'm so sick of these "price is dropping because of X" bullshit threads. Price is dropping because it had no business being $150+ in the first place. It's way too early for that kind of valuation compared to actual current use as currency. It was a BUBBLE, and bubbles always pop.

I'm sick of the "Price had no business being 150+" threads. What would be a fair price for a random unit of a currency backed by ... nothing?  No economy, no product, no army?  This is so frickin' arbitrary.