If You Could Build ASICs, Why Would You Sell Them?

[l008com]

I know this point has been raised before, in other threads. But I wanted to start a new thread about this point specifically. Regarding Butterfly Labs specifically, but also just in general. If they could make the devices they are "selling", why would they sell them? Even if they had to pay their own full MSRP for them, they should be able to simply run them and use them, and make way more money than they could ever make selling them! And keep in mind that full MSRP is likely at least 50% profit, if not more. Why would any ASIC maker at all, sell their boxes? It doesn't make sense to me, is there part of this I'm missing?

[1714037176]

1714037176

[1714037176]

1714037176

[1714037176]

1714037176

[anderl]

What makes you think they aren't doing that already.  Mine with the ASICs during the "test" phase.  When the difficulty skyrockets to the point that it is not profitable or they start seeing too many cancellations start shipping a few of the less profitable miners like the Jalapenos.  Difficulty goes up higher, sell the bigger models.

Then next year market the new 2500GH/s miner take another batch of preorders, rinse and repeat.

The reality is that it is about the business model.  Why don't gold miners mint their own coins or make their own jewelry.  Its not their business.  If you spread yourself thin you will have lower profit margins.  Better to focus on one thing and be the best in that market.

[HenryTheHorse]

Expectation seems greater to keep them to yourself.

[Foxpup]

Because if you keep your ASICs to yourself, you're betting that your competitors won't develop their own ASICs and drive the difficulty way up, because if they do, you suddenly can't make a profit anymore. But simply selling your ASICs immediately is a guaranteed way to make a profit, and the potential for a future difficulty rise is your customers' problem, not yours. Sure, there's a good chance you could make more money mining than by selling, but it's still only a chance, and when you spend millions on design and manufacturing equipment, you just can't afford to take chances.

[pcrsweetness]

yes good points are made and I'm sure they were using them before starting to ship. But as someone else pointed out, the mining manufacturers during the gold rush made more than the miners.  So is is probably just good business.

[Dabs]

In Avalon's case, they mentioned that no employee or owner of Avalon is running their own machines. They were tested to work but not in the bitcoin network, just the chips that they indeed can do the hashing. Now they are selling the next batch, and also the chips themselves.

Let's compute: One batch is 600 units * 75 BTC = 45,000 BTC.
10,000 chips = 708 BTC or 500,000 chips/10,000 = 50 * 708 = 35,400.

Actually, you can just look at the block chain for the payment addresses and you can see that they've altogether received in excess of 100,000 BTC. Which they have now. If they mined it, maybe they wouldn't get that much immediately.

Also, the owner or founder is a young idealist, who wants bitcoin mining to be decentralized, thus putting his units in the hands of many others.

[anderl]

Yes Avalon is a trusted company but for every good company there is an Enron that is out for themselves.

[bryanjones267]

if you sell 100 ASICS on a 1000$ price, you win more than mining with one that would give you 50.000 for example.

if you sell a lot of them, your clients will be competing among each other, but your profit is there.

p.s- you can always keep some to yourself, while selling others

[Balthozar]

A lot of the providers of the machines will tell you "We are "Stress testing" before shipment". So what do you think they are doing? They are running the shit out of the machines, making their mining profit, and then selling it to you at a point when the difficulty goes up. There was actually one company that was debating weather or not to release there Batch#3 or to keep them for them selves.

[bitsalame]

A lot of the providers of the machines will tell you "We are "Stress testing" before shipment". So what do you think they are doing? They are running the shit out of the machines, making their mining profit, and then selling it to you at a point when the difficulty goes up. There was actually one company that was debating weather or not to release there Batch#3 or to keep them for them selves.

Google bitcoin testnet

[downying]

it only makes sense to sell on auction

[Drizzle]

When the ASICs were designed, bitcoin price was a lot lower. I imagine that at the current prices it will be more tempting to run them, because they'd make more money in a week than selling them would. Then again, I have no idea how bitcoin profits would be taxed, which might affect things.

[coincepts]

the whole ASIC fiasco is questionable at best.

[ISAWHIM]

The thing is this...

You CAN NOT "estimate" the "returns" of a GH/s machine compared to a MH/s machine. In the end, the system "balances-out", leaving the GH/s machines making the same/less money than the MH/s machines.

How? (Glad you asked!)

Only X number of bitcoins are "allowed to be found" per time-period. EG, if you mine with a kH/s machine, you would find 1 per year (solo). If you mined with a MH/s machine, it would adjust until it only allowed 1 per year (solo). If you mined with a GH/s machine, it will adjust until it only found 1 per year... etc...

Now... The issue is this...

The "adjustment" takes a little time. The first hundred machines will reap, only for a few days, then the rates will change, and they will be exactly where we are now, with MH/s machines. But, they will be using less electricity, potentially. And..... AND.... all the MH/s machines will be making 1/1000 of what they make now, since, the bump-up value will HAVE TO DEGRADE by 1000, to compensate for the 1000x faster machines.

This will essentially kill all the WASTEFUL machines/rigs, making them unprofitable.

Here is the kicker...

If no more transactions are created, those GH/s machines will spend 90% of the time doing nothing. No work, not generating coins. Because they will finish all transactions and have nothing to hash. Transactions would have to be 1000x more, for them to "be equal to the MH/s machines now". Thus, they will actually make exactly the same as we make now, but it will just cut us out of the equation.

Why sell them? Because they know this. Also, because I am SURE, for every ONE they sell to us, they are building TWO for themselves. Thus, ensuring that WE never surpass them. If they are even building any for us at all. (I have only heard of a couple dozen being sent, for the millions they collected. Some still collecting, and have not sent anything in over a year. With lame excuses after excuses after excuses... even the chinese!)

In the end, generating MORE makes you LESS, because that degrades the value of the coins. If "Everyone" has Gh/s machines, and makes millions of coins... then the coins are worth less than $1.00 at the end of the day. (Which is why the system has a higher difficulty adjustment and a max-cap.)

The "failures" and "error-rates" and "stale" creations will rise also, naturally, as the end of the last minted coins come into creation. That will be about 90% wasted hashing, with 90% down-time due to "no one buying coins", because everyone is trying to "sell created coins", which they are trying to "sell for what they lost", and thus, "have no monetary value we can ever use."

By that time, the 512-hash Byte-Coin market will emerge, and make all these ASIC systems useless, since they only do 256-hashing... Then we crank out our old GPU's and start all over again! Let the vultures with billions in unsellable pennycoins, whine and cry about it, while we move on to new scams and confusingly over-technical money laundering methods that only hackers know how to use!

The worst thing these morons are doing is DDOSing their own lifesupport... That is like pulling the plug yourself.

[young3dvard]

I know this point has been raised before, in other threads. But I wanted to start a new thread about this point specifically. Regarding Butterfly Labs specifically, but also just in general. If they could make the devices they are "selling", why would they sell them? Even if they had to pay their own full MSRP for them, they should be able to simply run them and use them, and make way more money than they could ever make selling them! And keep in mind that full MSRP is likely at least 50% profit, if not more. Why would any ASIC maker at all, sell their boxes? It doesn't make sense to me, is there part of this I'm missing?

Developing ASIC is very costly, 100K USD and up, based on nm techology. Even if you have the money, you would need to be sure it will be profitale in 6-12 months after you start - very risky.

So the question is - why not risk other people money instead and have profit if all will go ok? seems a winner to me.

[🏰 TradeFortress 🏰]

BFL isn't doing testing on testnet through, as seen with Luke Jr's machine they're doing it on the main net.

[Dabs]

Transactions will continue, and probably even increase.

[Drizzle]

Only X number of bitcoins are "allowed to be found" per time-period. EG, if you mine with a kH/s machine, you would find 1 per year (solo). If you mined with a MH/s machine, it would adjust until it only allowed 1 per year (solo). If you mined with a GH/s machine, it will adjust until it only found 1 per year... etc...

Which is why it's best for those making ASICs to use them while they can make a decent amount of coins, and only sell them when the difficulty is such that it takes a couple of months to recoup the price of one.