ASICS killing BTC ?

[lucasjkr]

How do you complain about avarice when you have an ASIC, Asicminer shares, and a huge stash of LTC and PPC?

Simple. I ordered my ASIC out of greed nearly a year ago. Running the numbers and salivating just like everyone else. In the intervening period, seeing what's happened and extrapolating what will happen I think it's clear Asics are a net-negative to the currency. Sure, the network has many more hashes than it had a year ago. But that's not what's important. What it's starting to lack, and what will never come back, is the excitement that everyone had about bitcoin when it was something that everyone could participate in, to varying extents.

Put another way, bitcoin would be stronger, more valuable to everyone with 1/10 the network hashing power but 100x the miners, than this consolidation we have now. When the only investment required was a graphics card that many people already had, they participated and were excited about it. When the network devolves to exclude them, when its left with a few pillars of power (asicminer, a hand full of Avalon owners and a handful of mini rig owners), all of which require significant investment, the outside world will shrug their shoulders amd look on. As a consumer, for instance, I know that I have not been clamouring for an e-currency with irreversible transactions to send to vendors who may or may not be cooperative if there's an issue with their product. And I certainly don't wish for the means for my friends to send me money at no charge, only for myself to have to be hit with a wire fee if I need that money as cash. Especially not when they could just hand me cash or a check.

I very much doubt many other consumers wish for that either. But when they can at least participate, they then get excited about it and talk up those features. Don't believe me, that's fine. But that's my stance.  In the meantime, I'll continue on in my hashing, knowing or feeling like this is a dead end.

[NiteShdw]

Rant all you want, but it was inevitable.  Any time people see an opportunity to make money by getting ahead of the pack, they will, even if those profits are short lived.

Let me go back a few years.  I had an nVidia card that could only mine about 66 MH/s, which wasn't a long but I could earn a few coins a month, enough to offset the costs when coins were only $2-5. People then started buying 5970 and other cards and built mining specific rigs.  People like me that didn't believe at the time that spending a several hundred dollars on a dedicated miner was worth it, dropped out and stopped mining.  The mining started going to the guys building the dedicated GPU rigs with multiple cards, often buying from craigslist or ebay to snatch up as many cheap cards as they could.

Along comes ASIC, now you have people spending money upgrading to ASICs while those with the GPU mining can't compete and are getting pushed out of the market.

It's the same game.  I have an Avalon, but I know that within 6-9 months, the difficultly will be so high that I probably won't make enough to offset the cost of electricity.  Once I reach that point, my options are to either spend my saved up coins on a better, more efficient rig, or drop out.

So even though ASICs are taking over, a year from now, there will be Gen 2 ASICs that are twice as fast using half the power and will be much more profitable.  Join the crowd or get out of the way.

[DeathAndTaxes]

The largest ASIC player is owned by thousands of shareholders each who receive a relatively small amount of Bitcoins.  If Joe Smith average AsicMiner shareholder gets 3 BTC in dividends and sells them how is that any different than Joe mining 3 BTC and selling them.

No one average is getting 3 BTC in dividends. That is an early adopter big shot not the average Joe Smith. You probably don't own a single share.

Way to miss the point .... here so it won't offend your sensibilities ...  If Joe Smith average AsicMiner shareholder gets x BTC in dividends and sells them how is that any different than Joe mining x BTC and selling them.

As for me not personally owning any shares ... WTF does that have to do with the price of tea in China.

[felipelalli]

I agree that ASICS factories are bad to decentralization in short term, however the market is free and anyone can try do best anytime.

Crying is free too. In Portuguese we say "o choro é livre".

[Impaler]

I've come to agree with the OP's assessment, Dan Kaminsky agrees with us as well.  ASICs (and in fact every generation of mining improvement including GPUs) has a centralizing effect on mining and the effects are cumulative and each the generations have come too fast for any kind of re-democratization to occur.  And the current ASICs on the market now are not remotely the end of the line, theirs is a LOT more power that can be squeezed out of smaller feature chips.

Kaminsky thinks all this centralization weakens the network against regulatory pressure because the miners are so small in number but so individually BIG that they are worth targeting (theirs also collusion potential amongst miners).  I think that long before that happens the user base will vote with it's feet and move to what ever coin is more decentralized simply based on the threat that such a thing MIGHT happen.  Currently LTC is that obvious go to coin but it will probably get ASICs some day too (and soon too if it supplants BTC).

The long term solution is probably a chain that mixes a large number of hash algorithms together so as to make ASICs quite wasteful and expensive to develop as you would need a specially designed and fabricated chip for every algorithm.  Even then it might be necessary to defensively rotate in and out some algorithms if they are targeted for ASIC development.

[hate_the_face]

The largest ASIC player is owned by thousands of shareholders each who receive a relatively small amount of Bitcoins.  If Joe Smith average AsicMiner shareholder gets 3 BTC in dividends and sells them how is that any different than Joe mining 3 BTC and selling them.

No one average is getting 3 BTC in dividends. That is an early adopter big shot not the average Joe Smith. You probably don't own a single share.

Way to miss the point .... here so it won't offend your sensibilities ...  If Joe Smith average AsicMiner shareholder gets x BTC in dividends and sells them how is that any different than Joe mining x BTC and selling them.

As for me not personally owning any shares ... WTF does that have to do with the price of tea in China.

because Joe Smith the miner actually CREATED something, whether it is digital or physical, it still exists. Joe Smith the shareholder simply threw money at someone in the hope that they knew what they were doing and walked away. He isn't going to spend his coins in the slightest, just let him know when its time to cash out.

Joe the miner is the reason developers create services for coins because they know he will spend them

Joe the shareholder is the exact type of person who Bitcoin was designed to combat, a get rich quick suit who couldn't even spell Bitcoin if you gave him all three vowels

so why don't you stop worrying about offending other people's sensibilities and realize that by putting the CREATION process in the hands of few, it completely kills the whole decentralization argument.

Which is the only thing keeping Bitcoin from going belly up, it being too decentralized for world governments to try and attack.

But god damn, some of these fucking people are trying really hard to make that not true anymore

[Trongersoll]

I don't know, for roughly 1 BTC you too can be an ASIC Miner. I don't see a centralizing factor unless the little guys decide it just isn't worth it.

[Rassah]

Kaminsky thinks all this centralization weakens the network against regulatory pressure because the miners are so small in number but so individually BIG that they are worth targeting (theirs also collusion potential amongst miners).  I think that long before that happens the user base will vote with it's feet and move to what ever coin is more decentralized simply based on the threat that such a thing MIGHT happen.

I would hope that big miners who, unlike regular users, actually have a huge incentive to keep mining, would also be aware of this issue, and would base themselves in multiple countries, just to make sure they don't get taken down all at the same time. Plus I would hope that all the users will go to where all the hardware and software development is taking place, and all the hardware and software developers will go to where all the users are, and not just where all the mining takes place. (hardware as in ATMs, hardware wallets, etc) I.e. if this ASIC thing and centralization is really an issue, then people won't leave bitcoin for anything else until there is a catastrophic failure from this issue, which may turn people off from crypto-currencies completely for a long while.

[Cubic Earth]

I think SHA-256 is probably okay for proof of work.  Bitcoin, however, should include something other than just proof of work.  I am curious about the possibility of tying proof of work to proof of identity via private keys.  A block would have to have both a hash of a certain difficulty and be signed with a valid private key.  The hard-code of client could prohibit reject any block not signed by a valid private key, and to discourage centralization, only 4 out of the last 100 blocks could be signed by the same private key.  I just made up 4 out of 100... obviously these could be picked carefully after some statistical analysis.  We would first need to decide just how dispersed we wanted the hashing to be.

Then the real questions becomes 'what determines if a private key is valid?'.  Maybe those keys could be mined themselves, via scrypt.  Maybe it could be based on a web of trust model that somehow limited each person to just on current private key.  The private keys could wear out... they could be good for only 100 blocks total.

Of course we would have to be extremely cautions in picking a method to make sure it couldn't be easily gamed and lead to greater centralization.  There would be a market for valid private keys, and that is okay and expected.  I would hope there is someway the web of trust idea could be implemented.  Bitcoin, after all, is a service in the name of us people.  This project is not about trusting no one, but more so trusting everyone.  We trust ourselves - as a community - right now.  The web of trust could start with the core devs and valid keys would be propagated throughout the bitcoin community.  One person, one valid key.  The details of how that could be enforced I don't know yet.  But I hope some of you will be able to build on this.

[AliceWonder]

Let's say I'm a notorious hacking group or even a government that has backdoors into hundreds of thousands of person computers across the globe.

What's to stop me from using CPU mining to pull off a 51% attack?

ASICs - that's what.

With CPU/GPU mining obsolete, that huge network of compromised machines is worthless for a 51% attack.

[Entropy-uc]

How do you complain about avarice when you have an ASIC, Asicminer shares, and a huge stash of LTC and PPC?

Simple. I ordered my ASIC out of greed nearly a year ago. Running the numbers and salivating just like everyone else. In the intervening period, seeing what's happened and extrapolating what will happen I think it's clear Asics are a net-negative to the currency. Sure, the network has many more hashes than it had a year ago. But that's not what's important. What it's starting to lack, and what will never come back, is the excitement that everyone had about bitcoin when it was something that everyone could participate in, to varying extents.

Put another way, bitcoin would be stronger, more valuable to everyone with 1/10 the network hashing power but 100x the miners, than this consolidation we have now. When the only investment required was a graphics card that many people already had, they participated and were excited about it. When the network devolves to exclude them, when its left with a few pillars of power (asicminer, a hand full of Avalon owners and a handful of mini rig owners), all of which require significant investment, the outside world will shrug their shoulders amd look on. As a consumer, for instance, I know that I have not been clamouring for an e-currency with irreversible transactions to send to vendors who may or may not be cooperative if there's an issue with their product. And I certainly don't wish for the means for my friends to send me money at no charge, only for myself to have to be hit with a wire fee if I need that money as cash. Especially not when they could just hand me cash or a check.

I very much doubt many other consumers wish for that either. But when they can at least participate, they then get excited about it and talk up those features. Don't believe me, that's fine. But that's my stance.  In the meantime, I'll continue on in my hashing, knowing or feeling like this is a dead end.

This is exactly what I feared when the BFL disaster began.

A fool gets conned by a group of smooth talking hucksters and pays for a golden goose laying ASIC device.  There is plenty of evidence that the hucksters have no idea how to build such a device, but greed parts the fool and his money.  Enough fools throw money at the hucksters that some real players take note and start developing their own devices.  The hucksters promise '2 weeks' every few days, and that keeps the fool dreaming of sugar plums rather than investing in alternatives.

When the real players deliver and the hucksters still can get their shoes tied, the fool's dreams of sugar plums fade slowly.  Eventually difficulty rises so much that the fool realizes gifting his money to hucksters was a huge error, and he starts counting all he he lost by trusting the hucksters.

Who does the fool blame?  Not himself, not the hucksters, he blames Bitcoin!

[lucasjkr]

What are you talking about? You're putting words in my mouth, or not at all understanding what I said....

 I'm doing fine. my jalapeno cost $149 - in 7 days, I've generated just about 1 coin. I'm FAIRLY positive that it'll generate another coin at some point. So, I'm not blaming ASIC's at all on my "loss" (which hasn't occured). I'm blaming ASIC's for driving many people away from Bitcoin. I'd hardly call myself "conned". Sure, it took a lot longer than expected, but I'm certainly glad I declined my friends offer to split the cost of several GPU's back in May to run while I waited for this.... That would have been a disaster. But no, I'm not part of the problem, driving away ordinary peoples interest in Bitcoin, I fully acknowledge that.

But still, hear me out... Bitcoin without people to drive interest in it isn't going to amount to anything. If you think that network hash rate is more important than network participants, well, I would be shocked.

[hate_the_face]

What are you talking about? You're putting words in my mouth, or not at all understanding what I said....

 I'm doing fine. my jalapeno cost $149 - in 7 days, I've generated just about 1 coin. I'm FAIRLY positive that it'll generate another coin at some point. So, I'm not blaming ASIC's at all on my "loss" (which hasn't occured). I'm blaming ASIC's for driving many people away from Bitcoin. I'd hardly call myself "conned". Sure, it took a lot longer than expected, but I'm certainly glad I declined my friends offer to split the cost of several GPU's back in May to run while I waited for this.... That would have been a disaster. But no, I'm not part of the problem, driving away ordinary peoples interest in Bitcoin, I fully acknowledge that.

But still, hear me out... Bitcoin without people to drive interest in it isn't going to amount to anything. If you think that network hash rate is more important than network participants, well, I would be shocked.

tell that to Rassah, dude is reminding me of Denzell in Training Day when Ethan Hawke walks away from him

http://www.youtube.com/watch?v=6KrNpxODiDA

people are walking away Rassah, DEAL WITH IT OR ADAPT

[pallokala]

ASICS are not killing Bitcoin, they are moving it into the future.

This. ASICS make mining harder for new people joining Bitcoin, but they also provide much more support for blockchain.

[bigbeninlondon]

Hey, I've been wondering, Litecoins are not immune to ASIC mining, they just need ASICs with a whole lot of memory to do it, meaning they would be way more complex and expensive. Once Litecoin goes up in value to the point where it would be worth it to invest in Litecoin ASICs, won't Litecoin become even more centralized, just because Litecoin ASICs will be very expensive to own and owned by a select few wealthy types, compared to Bitcoin, for which ASICs are dirt cheap (after initial design expense)? Right now it seems like Litecoins are more decentralized, but I fear eventually Bitcoin miners to Litecoin miners will be like Honda owners to Ferrari owners.

Memory is a lot cheaper than people seem to realize.  Slap some DDR3 sockets on your ASIC board and you have 16gb for like $60.  I understand that memory is more expensive than computation power, but not forever, and not nearly to the degree everyone here seems to think.  The reason ASICs don't exist for Litecoin right now is because no one sees the same profit margins because adoption is so much lower.  Litecoin is based almost PURELY on speculation, as opposed to bitcoin that actually has a small viable market.  In addition, Litecoin is mainly mined by people who like mining but can't mine bitcoin at a profit any longer.

[bigbeninlondon]

So, people who invested heavily in mining equipment with hopes of getting rich quick are causing a surplus of bitcoin supply, thus driving the price down and selling their newly minted bitcoin to people who want to hold on to/spend bitcoin at a reduced rate.

I don't see anyone losing here or anything being killed, except people who overvalued coins when they were hot, and weren't planning to hold them long-term are getting the shaft. That's the inherent risk in any market.

The number of coins created remains constant; that's what difficulty is for.  So there is indeed no surplus of bitcoin supply.

[lucasjkr]

ASICS are not killing Bitcoin, they are moving it into the future.

This. ASICS make mining harder for new people joining Bitcoin, but they also provide much more support for blockchain.

They barely provide any protection. Well, they protect it from script kiddies, but even with tons of GPU's, there wasn't too much of a worry there... after all, 99% of them are looking to make money, so all theyed do is change someones payout address, which they can do whether a machine is CPU mining, GPU mining or ASIC mining.

For an attacker intent on taking down bitcoin, they provide next to no protection... First, that attacker isn't looking for profit. Second, by contemplating attacking bitcoin, they're well aware that it's going to cost 7 or 8 figures. And a 7 or 8 figure expense is a rounding error in the budget of most governments or banks, any of whom could contract the developement of their own ASIC and still take it down with relative ease.

[Littleshop]

ASICS are not killing Bitcoin, they are moving it into the future.

This. ASICS make mining harder for new people joining Bitcoin, but they also provide much more support for blockchain.

Nope.  Litecoin now has more protection then bitcoin.   Compromise two large hosts now (asic mining operation and large pool) and you have 51%+.  ASICs are good for bitcoin if they are widely distributed.  In the hands of a few they are a problem. 

[mnyonpa]

I am sorry if I repeat something already been said on the thread, or if I say something stupid because I am new to Bitcoin business, but I see no danger of ASICs killing Bitcoin, and I’d be happy if you please show me where exactly I am mistaken.

If I understand correctly, there are hobbyist miners and professional miners. Hobbyist miners (like myself) engage in mining out of sheer curiosity and their ultimate desire is to get some extra bitcoins to spend on a pizza. A hobbyist miner does not invest considerable money into mining nor awaits a quick ROI and considerable profit. Professional miners are those organising mining rigs and investing serious money into the process of mining. Professional miners crave for serious mining power in order to get profit.

Professional miners did not appear with ASICs. I guess they started to emerge when GPU mining replaced CPU mining. I doubt that someone would buy a couple of computers solely for mining purposes, but GPUs are much cheaper and you can stuff a whole lot of them together with the cooling making a powerful mining rig. Most of the mining power of the network since the appearance of rigs and later FPGAs began to concentrate in the hands of professionals. Did it kill bitcoin? Obviously, no. Do you really think that 1000 hobbyist miners who mine on their only GPU brought any considerable power into the network? Just 10 professional miners with a rig of 100 GPUs each can overpower those. Make 80% of the hobbyists go but the network will stay. And the number of hobbyists won’t dwindle because ASICs will only become cheaper. On the contrary — one cannot stick 10 GPUs into a cheap netbook, but can easily do the same with a pile of USB Erupters, and something like Erupter Blade does not require a computer at all!

Now there come ASICs. Does it considerably change anything about the division between hobbyist miners and professional miners? No. The former instead of wasting electricity and killing their GPU can buy themselves some USB Erupters, or several months later even a Jalapeño or analog. Environment is safer, everyone is happier. The latter exchange their GPU rigs for ASIC rigs. The early adopters have more competing power, just as you do in business. Weaker ones get out. Stronger ones find their niche with an amount of Gh/s they can afford to invest in. Nothing really changes much. Again, the environment is safer.

But mining is only one faucet of the Bitcoin network and it’s not the most import one! If everything revolved around mining, Bitcoin would never get any value. The real value goes from adopting Bitcoin in businesses. And I believe, with ASICs the network becomes more protected against attackers because it is really hard to accumulate 51% of current TH/s and PH/s that are coming. With more protection and possibly new improvements to the protocol we have more adopters, and the value of Bitcoin rises!

The only real problem are governvents prohibiting Bitcoin collectively. And no government will do that through ASICs or other hackery. It is much simpler. Declare Bitcoin illegal. Close all exchanges into local fiat — and you will constrict Bitcoin to underground barter economy only. Can this happen? Theoretically yes, but very unlikely. Even if the US or the EU expels Bitcoin into underground, we have a growing adoption in China and Latin America. And for some states Bitcoin is not a danger to their economy, but just the opposite — a way to get less dependent on the first world banking system.

The only real danger to Bitcoin is if most of the adopters will be speculators and high volatility drives real producers out of Bitcoin. But the ASIC hype actually targeted speculators and those looking for get rich quickly schemes. Now when a considerable part of them leaves due to the stall with new ASICs and drop of Bitcoin value this will only improve the new more secure Bitcoin network.

[DeathAndTaxes]

They barely provide any protection. Well, they protect it from script kiddies, but even with tons of GPU's, there wasn't too much of a worry there... after all, 99% of them are looking to make money, so all theyed do is change someones payout address, which they can do whether a machine is CPU mining, GPU mining or ASIC mining.

For an attacker intent on taking down bitcoin, they provide next to no protection... First, that attacker isn't looking for profit. Second, by contemplating attacking bitcoin, they're well aware that it's going to cost 7 or 8 figures. And a 7 or 8 figure expense is a rounding error in the budget of most governments or banks, any of whom could contract the developement of their own ASIC and still take it down with relative ease.

Well no.  There is no bank where 8 figures is a "rounding error" and most nations could not afford that type of expenditure.  Still Bitcoin is hardly worth spending 8 figures to kill it today however someday Bitcoins might have 100x the economic activity and then spending 10 figures is beyond a rounding error for just about everyone on the planet.

The cost in killing Bitcoin is that the second you do someone will release a newer coin that is immune (or at least very resistant) to the method you used to kill it.  Think POW/POS hybrid or something we haven't even thought of yet.  Necessity is the mother of all invention.  Much like killing Napster didn't kill file sharing.  So what is next spend another 8-12 figures killing the half dozen next gen hybrids spawned from the carcass of Bitcoin?  Then what spend another 8-12 figures killing their hybrids.  It is merely forced evolution.   

ASICs does three things:
a) prevents attackers from "cheating".  Pretend ASICs don't exist and the attacker uses a technology over a magnitude more efficient.  Bitcoin GPU miners march into battle with flintlock rifles and get nuked from orbit by a star destroyer. 

b) finally eliminates the risk of botnets.  Some people think GPU have done that but the on core GPU (APU) have significantly improved and now Intel and AMD both include OpenCL drivers in their default installs.  It is only a matter of time before botnets start adapting to use GPU power.  Now the on-core GPU may seem but they are capable of 50 to 100 MH/s.  A botnet with  ten thousand such nodes would have significant power against an "all GPU" network.

c) has the potential of creating widespread hashing power on a scale never see before.  Think $5 chip in home router (Bitcoin edition) adding 1 GH/s to the network.  Now imagine someone like dlink is making the router and sells 50 million units.