Cryptocoin Mining?

[hawtdawg]

One of those systems is 1 terahashes/sec! These are very risky investments though since GPUs have other uses whereas these ASIC miners don't.

the 1300 dollar one, and the USB one aren't too risky really. If you can get one at launch, they'll likely pay themselves off pretty fast.

 

[RussianSensation]

Butterfly labs ASIC miners will make GPU mining completely obsolete here in about 3-4 months.

http://www.butterflylabs.com/products/

look at the stats/prices for the Bitforce SC miners. GPU mining is officially dead the moment these hit the streets. I've got 2 of the 40 GH/s ones on preorder and am probably about to start trying to get rid of my 6950's.

That sounds almost unreal. That would make $5500 / month @ $8 rate. Doesn't make any sense. How did they just go from 832 Mhash/sec for $600 and 25 GH/sec for $15,000 to 40 GH/sec for $1300?

If true, I'll get a couple, but still I'll wait to see if that performance is actually real!

Right now BTC Guild's entire pool speed is 1400-1500 GH/sec while Slush's Pool is 1100-1300 GH/sec. You realize if 100s of these units enter the market, the network difficulty would increase so fast as a result of ASIC market penetration that it would most likely finish the entire Bitcoin mining operation. We'll have difficulty increasing every day instead of months.

 

[thujone]

well... this makes my decision on whether or not i should get a ati over nvidia for bitcoin mining a whole lot easier.


given those are coming out soon... why bother? lol

 

[pcm81]

The real question is this:
At current difficulty/price you can pay off the 40GHps system in 1 week. Why would the bother selling them instead of running them?

 

[RussianSensation]

well... this makes my decision on whether or not i should get a ati over nvidia for bitcoin mining a whole lot easier.


given those are coming out soon... why bother? lol

It's mid-July, and those units aren't coming out until October. Let's assume October 1st. That leaves 2.5 months of mining. A single 7970 makes about 11 BTC a month at 1150mhz+. Current rate is > $8!

That's 11 BTC x 2.5 months x $8 = $220. Still way cheaper than a GTX670/680, not to mention there is no trade-off in gaming performance.

These boxes have their uses out side of mining as well, it is just less accessible to the layman. Hashing power has applications outside bitcoin and password cracking.

"No, their function is limited to high speed encryption validation in the specific double step sha256 protocol. It’s not useful for any purpose related to rainbow tables or password recovery."
http://www.butterflylabs.com/bitforce-sc-faq/

 

[taltamir]

Yep. It was only a matter of time before ASICs ruined GPU mining, same way GPU mining destroyed CPU mining. The difference is, I think the period where ASIC mining will be very profitable will be shorter than the period when GPU mining was very profitable. GPU mining was quite profitable from November 2010 till September/October 2011. I don't think ASICs will enjoy the same profitability period as GPUs did.

They are selling people a money manufacturing device that anyone can use with 0 training and skill; and has no purpose other then this manufacturing. Just buy it and turn it on and watch the money come in.

The only reason you would sell something like that rather then just run 1000 of them in a warehouse is because you hope to over-saturate the market, selling so many the market crashes and every one of your customers loses money, making less then the cost of the product.

If they are REALLY wicked they could run a bunch of those to eke out the profit from the market before it collapses before actually shipping the product to preorderers.

 

[3DVagabond]

The only reason you would sell something like that rather then just run 1000 of them in a warehouse is because you hope to over-saturate the market, selling so many the market crashes and every one of your customers loses money, making less then the cost of the product.

The simple, less devious, reason is that the hardware is more profitable to sell than to run.

 

[f1sherman]

and the each one they plug into network lowers it's profitability

 

[pcm81]

But, if hardware is more profitable to sell then to run, then it's not worth buying...

 

[taltamir]

The simple, less devious, reason is that the hardware is more profitable to sell than to run.

That is what I said.

The problem is that there is only ONE possible way for it to be cheaper to sell the hardware then to run it. And I explained what it is.

f1sherman and pcm81 get it though

 

[RussianSensation]

Honestly if these new Butterfly Labs mining systems are real, the entire bitcoin operation will be over.

This $30k Mini Rig is supposed to do 1000 GH/sec:
http://www.butterflylabs.com/order-form-bitforce-sc-mini-rig/

At that rate @ $8 per BTC, it would mine 17,000 BTC in 1 month or > $130K (i.e., break even in less than 1 week). The difficulty level would 'hockey stick' if just 10 of these join the network, completely destroying mining for 99.9999% of the market. But the only reason any currency has value is if it is used to buy/sell items. If 99.9999% of the market cannot participate in market transactions involving the said currency because they have no bitcoins, the value of bitcoin should be basically worthless because the active transaction market for bitcoins would cease to exist. Also, the few individuals remaining with a large bitcoin pile will be hoarding it or unable to sell due to lack of interested parties involved in transactions as a whole. Maybe I am missing something?

 

[3DVagabond]

That is what I said.

The problem is that there is only ONE possible way for it to be cheaper to sell the hardware then to run it. And I explained what it is.

f1sherman and pcm81 get it though

Why does Mac Trucks sell trucks? By your reckoning they'd make more money is they just drove them themselves.

 

[pcm81]

Honestly if these new Butterfly Labs mining systems are real, the entire bitcoin operation will be over.

This $30k Mini Rig is supposed to do 1000 GH/sec:
http://www.butterflylabs.com/order-form-bitforce-sc-mini-rig/

At that rate @ $8 per BTC, it would mine 17,000 BTC in 1 month or > $130K (i.e., break even in less than 1 week). The difficulty level would 'hockey stick' if just 10 of these join the network, completely destroying mining for 99.9999% of the market. But the only reason any currency has value is if it is used to buy/sell items. If 99.9999% of the market cannot participate in market transactions involving the said currency because they have no bitcoins, the value of bitcoin should be basically worthless because the active transaction market for bitcoins would cease to exist. Also, the few individuals remaining with a large bitcoin pile will be hoarding it or unable to sell due to lack of interested parties involved in transactions as a whole. Maybe I am missing something?

this is completely wrong... Lets say the network hash rate went up by a actor of 1000. Difficulty goes up by a factor of 1000. Your 6990 miner now is making 1000 times less bitcoins per day. BUT the real question is this: did the value of bitcoin scale with the difficulty and now each bitcoin costs $8,000? If answer is yes, then all is well.

 

[taltamir]

Why does Mac Trucks sell trucks? By your reckoning they'd make more money is they just drove them themselves.

no, trucking requires customers. A truck manufacturer will never manage to round up enough customers themselves but if they sell their trucks, each buyer gets customers and pays for their investment.
This applies for every tool of industry.

But a dedicated bitcoin miner is not a tool of industry.
You don't need skill, you don't have customers, you don't worry about sales...
You plug it in and it generates money (literally!) out of thin air.

At that rate @ $8 per BTC, it would mine 17,000 BTC in 1 month or > $130K (i.e., break even in less than 1 week). The difficulty level would 'hockey stick' if just 10 of these join the network, completely destroying mining for 99.9999% of the market. But the only reason any currency has value is if it is used to buy/sell items. If 99.9999% of the market cannot participate in market transactions involving the said currency because they have no bitcoins, the value of bitcoin should be basically worthless because the active transaction market for bitcoins would cease to exist. Also, the few individuals remaining with a large bitcoin pile will be hoarding it or unable to sell due to lack of interested parties involved in transactions as a whole. Maybe I am missing something?

maybe they figure that if they use it themselves the coins would be worthless due to what you described so by selling it they let every owner of one of their systems participate.

However, that still collapses the market from "everyone with a GPU willing to mine" to "everyone who paid a minimum of 500$ to this one company". And the coins will never pay for themselves. And the sheer efficiency of those devices is such that they will cause hyper inflation overnight permanently ruining public trust in the currency.

Either they are completely clueless about the most basic aspects of economics or they are malicious and want to destroy bitcoins.

 

[LightRider]

Now we see the invalid nature of the monetary system come to light. Why use money when its manufacture and upkeep do nothing useful for society?

 

[taltamir]

Now we see the invalid nature of the monetary system come to light. Why use money when its manufacture and upkeep do nothing useful for society?

They actually contribute greatly to society. Barter is a very inefficient and difficult to handle system. You cannot have standardized pricing and/or carry large quantities of stuff.

Something will be used as money naturally though. Something precious, unlikely to spoil (unlike potatos), with a standard relatively stable-ish value. Things like salt or gold or oil.

The problem with a currency which is essentially a precious good (gold, silver, oil, salt) lies in rapid loss of value when a new mining/manufacturing method is discovered (salt is worthless today).

Which is not nearly as bad as the problem with fiat currency is its mismanagement by the idiots in charge of countries that result in countries collapsing. However this is maintained due to the power fiat currency gives and as such it is enforced by those with a monopoly on violence via violence.

But since you can't change it why worry about it. Just do the best you can with the fiat money system and take into account the yearly losses due to government. As well as trading to more wisely managed currencies.

 

[Chiropteran]

Honestly if these new Butterfly Labs mining systems are real, the entire bitcoin operation will be over.

This $30k Mini Rig is supposed to do 1000 GH/sec:
http://www.butterflylabs.com/order-form-bitforce-sc-mini-rig/


...


Maybe I am missing something?

It's vaporware at this point. People were saying the same things about the original mini rig and single but they ended up with actual specs much lower than the original rumored values.

 

[hawtdawg]

It's vaporware at this point. People were saying the same things about the original mini rig and single but they ended up with actual specs much lower than the original rumored values.

if they're 1/4 of what they're claiming, it will still completely kill GPU mining.

 

[Chiropteran]

if they're 1/4 of what they're claiming, it will still completely kill GPU mining.

It will also completely kill all other products sold by butterfly labs. Why would they release it and make all their other products obsolete? I doubt it'll ever exist as described.

 

[hawtdawg]

It will also completely kill all other products sold by butterfly labs. Why would they release it and make all their other products obsolete? I doubt it'll ever exist as described.

That's like saying that it would have been stupid for AMD to release the 7970 because it made the 6970 obsolete.

 

[Chiropteran]

That's like saying that it would have been stupid for AMD to release the 7970 because it made the 6970 obsolete.

No it's not. It's like saying if people told me the 7970 was going to be 30X as fast as the 6970 but it's only going to cost 50% more, I'd take it with a grain of salt. There is normal progress, and there is exaggeration.

I just don't think 1000ghash for 30k is very realistic at all.

Why would BFL do that? They are selling out and constrained by their supply on selling singles already, whatever they release that is better will almost certainly sell out just as fast and quick, they are only going to be limited by the manufacturing on their side. If the current stuff is selling out so fast, why would they release a new product with the price per mhash reduced so dramatically?

Release something with double the performance per dollar, that would make sense. But these numbers do not.


edit: That said, if it is true there will be some major bitcoin deflation. As posters above suggested, if you increase the total hash power by 10X the value of each bitcoin needs to also increase 10X to keep up. If you honestly believe the specs are 100% real you should go out and buy buy buy all the bitcoins you can afford.

 

[taltamir]

No it's not. It's like saying if people told me the 7970 was going to be 30X as fast as the 6970 but it's only going to cost 50% more, I'd take it with a grain of salt. There is normal progress, and there is exaggeration.

I just don't think 1000ghash for 30k is very realistic at all.

Why would BFL do that? They are selling out and constrained by their supply on selling singles already, whatever they release that is better will almost certainly sell out just as fast and quick, they are only going to be limited by the manufacturing on their side. If the current stuff is selling out so fast, why would they release a new product with the price per mhash reduced so dramatically?

Release something with double the performance per dollar, that would make sense. But these numbers do not.

Yep

edit: That said, if it is true there will be some major bitcoin deflation. As posters above suggested, if you increase the total hash power by 10X the value of each bitcoin needs to also increase 10X to keep up. If you honestly believe the specs are 100% real you should go out and buy buy buy all the bitcoins you can afford.

You mean price INFLATION and you should SELL.
If quantity of coins is increased 100x their value will drop 100x and owning a bitcoin is a bad idea when that happens. Price is then inflated, what cost 1BC now has inflated to cost 100BC.

There is one thing that occurred to me. Bitcoins are meant to be an alternative to fiat curruncies of the world with the precious good in question being rare, difficult, unique, math solutions. They operate under the delusion that it cannot be shut down by governments because its distributed.

It is possible that they are selling it not because its more profitable to sell them then to run them. But because they are afraid (justifiably based on history of every other alternative currency) that if they ran them all themselves they will be shut down, hard, by local government. This is actually a reasonable scenario.
Furthermore, if they ran it all themselves they would end up with 90% of the "currency" in their own hands and it will be much worse for the future of bitcoin as a viable currency then cutting out all GPU crunchers as has been predicted earlier in this thread.

And the issue with the market crash... it is bound to occur sooner or later and is a one time thing.
There was the CPU to GPU transition and now the GPU to ASIC transition. Someone is gonna do it and while it will be a setback, the currency might yet survive the crash (might, many BC institutions will not) and then they will be the sole suppliers for mining hardware. And you cannot have a repeat of that crash since there is no huge jump from ASIC

However, if they were smarter they would have provided 2x GHash /$ compared to current machinery at a fraction of manufacturing cost and power consumption. And every year repeated the doubling of the HHash (without even needing to develop new hardware)... if a competitor comes out with a similar product then they release their big guns. It would have resulted in getting paid more and also the gradual transition would have been less harmful to the market. Since the product IS money.

However, it is possible that they are intentionally going for broke believing in a "clean break"... and that the market will survive and recover and things will only get better then.

 

[kalrith]

You mean price INFLATION and you should SELL.
If quantity of coins is increased 100x their value will drop 100x and owning a bitcoin is a bad idea when that happens. Price is then inflated, what cost 1BC now has inflated to cost 100BC.

If efficiency increases 10-fold, then wouldn't the difficulty of mining and not the quantity of bitcoins increase? If that's the case, then I would assume the value of BC would increase.

 

[Chiropteran]

You mean price INFLATION and you should SELL.
If quantity of coins is increased 100x their value will drop 100x and owning a bitcoin is a bad idea when that happens. Price is then inflated, what cost 1BC now has inflated to cost 100BC.

No I meant deflation. Look at the historical value of bitcoins in relation to difficulty, as difficulty goes up value goes up.

Quantity will not increase by 100x, not even close. Until we go through a few difficulty adjustments we might have some faster blocks, and output of bitcoins might be increased a bit for a week or so, but overall it's just a drop in the bucket compared to all the bitcoins already in circulation.

After that brief increase in bitcoin output the difficulty adjustments will kick in and output will be about the same as it ever was, given the new levels of mhash. And when that happens, GPU mining becomes very inefficient.

ASIC mining becomes the new standard, the market will continue to adjust until it reaches a point where the new ASIC miners will pay off only after 6 months or even over a year of mining, as that has been the stabilization point for GPU mining. Anything less and there is just too much pressure to buy ASIC hardware, which in turn pushes difficulty up for everyone. Much higher, and the chance of ever making a return on your money starts to look slim and miners start quitting.

That said, this whole series of events is dependent on BFL's stuff being non-vaporware. Even if the hardware is released to match the specs we have now, supply will limit things a bit, and it might take a year or more before the market fully adjusts. There will be some rubber-banding even, as ASIC start to take over some major GPU miners might shut down in droves and temporarily cause a difficulty reduction, but over time the cheap ASIC will take over.

In the end though, mining will remaining profitable. To do that, there is going to be some deflation of the currency.

 

[thujone]

well... i decided to hop on that cheap 7950 deal and will hopefully be mining by the weekend.


i figure even if i do watch everything collapse... that'll be just about as entertaining as making money lol