ASICMINER: Entering the Future of ASIC Mining by Inventing It

[velacreations]

if you have a device that uses 28nm chips that is priced at $5/gh and another company is selling devices using 45nm priced at $3/gh, what makes you think the market won't buy the 45nm devices?

At this point in the mining game chip tech takes a back seat to cost.

[BitThink]

if you have a device that uses 28nm chips that is priced at $5/gh and another company is selling devices using 45nm priced at $3/gh, what makes you think the market won't buy the 45nm devices?

At this point in the mining game chip tech takes a back seat to cost.

The watt/G also plays a role. Suppose the 28nm takes 1w/g and 45nm 2w/g and I have a data center with 300kw power supply. With 28nm I can hash at 300T but only 150T with 45nm. So price is important but power efficiency also affects my choice.

[velacreations]

The watt/G also plays a role. Suppose the 28nm takes 1w/g and 45nm 2w/g and I have a data center with 300kw power supply. With 28nm I can hash at 300T but only 150T with 45nm. So price is important but power efficiency also affects my choice.

it depends on what sector of the market you are talking about.  W/gh has not been a major factor for small miners (under 10 TH) so far.  It may become more important in the future, but right now, cost/gh is significantly more important.

Basically, older tech can co-exist with newer tech, as long as they give a price advantage to certain sectors of the market.

[Endlessa]

if you have a device that uses 28nm chips that is priced at $5/gh and another company is selling devices using 45nm priced at $3/gh, what makes you think the market won't buy the 45nm devices?

At this point in the mining game chip tech takes a back seat to cost.

The smaller chip will always have a lower production cost per gh. . .and will have lower power consumption (which will become relevant sometime in feb).  After the manufacturer covers their setup costs, which probably happens in the first month or so with the way the current hardware frenzy is going,  it will be easy for them under cut the larger chips.   Think about it this way. .. .


I can put 1 chip on a board that does 250 gh/s per chip with a small die. or I can put 100 2.5 gh/s chips on a board and get 250 gh/s.  the reduction in the size of PCB, supporting hardware, assembly time/labor, power supply, cooling fans, etc.   will make the smaller single chip have much lower costs to produce than the larger one.  In the short term the larger chip will have a faster time to market and cheaper setup costs, but the smaller chip will compete better in the long term game.  Now, in the past AM was first to market, so they got good leverage on the larger chips, but now smaller chips than their current target are already shipping, so that advantage has already been lost.  Going to market with a larger chip is extremely risky, especially since this industry is all about power consumption and GH density.  . . imo

Hope that helps explain why

[velacreations]

this industry is all about power consumption and GH density.  . .

if that was the case, no one would have ever bought a USB miner.

[Endlessa]

this industry is all about power consumption and GH density.  . .

if that was the case, no one would have ever bought a USB miner.

Not necessarily true, the usb miner was attractive for people for several reasons.  The product was in-hand, relatively cheap, allowed people to experiment with little risk and there was a frenzy to get some kind of hardware.  It was the only thing available without long pre-order wait times.  That window has closed/is closing now.  There is going to need to be more competitive reasons for purchase in the future.

[velacreations]

Not necessarily true, the usb miner was attractive for people for several reasons.  The product was in-hand, relatively cheap, allowed people to experiment with little risk and there was a frenzy to get some kind of hardware.  It was the only thing available without long pre-order wait times.  That window has closed/is closing now.  There is going to need to be more competitive reasons for purchase in the future.

cost/gh will be/is the driving factor for a significant portion of the market.

[Endlessa]

Not necessarily true, the usb miner was attractive for people for several reasons.  The product was in-hand, relatively cheap, allowed people to experiment with little risk and there was a frenzy to get some kind of hardware.  It was the only thing available without long pre-order wait times.  That window has closed/is closing now.  There is going to need to be more competitive reasons for purchase in the future.

cost/gh will be/is the driving factor for a significant portion of the market.

correct

[BitThink]

Not necessarily true, the usb miner was attractive for people for several reasons.  The product was in-hand, relatively cheap, allowed people to experiment with little risk and there was a frenzy to get some kind of hardware.  It was the only thing available without long pre-order wait times.  That window has closed/is closing now.  There is going to need to be more competitive reasons for purchase in the future.

cost/gh will be/is the driving factor for a significant portion of the market.

The problem now is the USB price has to be lowered down several times and still on stock but the blader sold out quickly. Apparently the majority revenue and profit comes from the serious buyers, and this becomes more and more true after the difficulty increases further.  For them, power efficiency and hash rate density is essencial.

[Luckybit]

Really risky to buy in when there's no solid info about the Gen2 chips.

Moreover, before buying, make sure you have considered the risk that Cointerra may be successful in their 22nm chips and sell hardware at $3/GH on Jan 2014. I believe AM will survive next year, but not likely to have the same advantage as this year. You could look at AM's price at early this year when they were clearly the first ASIC miner, and then predict its price when they don't have advantage anymore now.

high return involves high risk, which is always the case.

It is true that there is no positive information about the gen2 chips. All we know is friedcat and his team are working hard on that. Based on what they have achieved this year, pay 0.9BTC/share to buy/risk/gamble/invest(whatever you call it) its success, is a good deal or not? You have to make your own choice.

It's difficult imo for AM to compete and recoop the dev costs of a new production line for several reasons:

* The smaller the chip size the larger the R&D/setup costs for production
* The currently expressed chip size will be undesirable to the market given that better alternatives are currently shipping
* The previous customer base hasn't/will not be re-cooping the cost of their previous expenditures on AM products and will not have much desire to spend on more
* If they do produce the expressed target chip size they will be at a cost disadvantage in what is soon to be a very tight profit margin squeeze for the industry

So then we have mining. . . AM has shown a propensity to not be able deploy and keep up with the network hashrate. aside from the beginning when it was easy, I have not seen them show the capability of reaching their target goals on hashrate.  Each time,  the target mining hashrate was met hugely late or missed completely.

Now, you can get all butt hurt and tell yourself I'm trolling, and I couldn't care less, but anybody new coming into this needs realize these shares should be entered into with trepidation and are not necessarily "discounted" for unfounded reasons.

g' luck all and happy trading

Hashrate wont continue to rise exponentially forever. At some point it will taper off and the share prices will go up again.

[SOSLOVE868]

Hashrate wont continue to rise exponentially forever. At some point it will taper off and the share prices will go up again.

I believe that Chinese company will at least get some portion in this game ...and I believe AM should be the one ....once that fully custom 55nm come out ,thus 28nm altera based system should be no way to compete with ,due to the cost of manufacture....

[romerun]

to da mooooooooon !

[ninjarobot]

I think you either need to be the best in $/GHs or in W/GHs - Both are viable strategies for now and different product lines can exist concurrently.

[VolanicEruptor]

I think you either need to be the best in $/GHs or in W/GHs - Both are viable strategies for now and different product lines can exist concurrently.

Watts per Gigahash is of much smaller importance than Cost per GH.  You are talking cents on the dollar and only becomes important near the end of the units life when it is struggling to stay profitable.  The beginning of its life is going to show far more profit than the end of it so of course cost per gh is most crucial

[pand70]

Hashrate wont continue to rise exponentially forever. At some point it will taper off and the share prices will go up again.

And what exactly is it that will stop hashrate from rising exponentially?

[tinus42]

I think you either need to be the best in $/GHs or in W/GHs - Both are viable strategies for now and different product lines can exist concurrently.

Watts per Gigahash is of much smaller importance than Cost per GH.  You are talking cents on the dollar and only becomes important near the end of the units life when it is struggling to stay profitable.  The beginning of its life is going to show far more profit than the end of it so of course cost per gh is most crucial

I don't care about the power costs. It's getting colder now where I live and will likely be cold for many months until next year's spring. All the power used by my mining gear will increase the heat in my house, resulting in a lower heating bill, which will offset the power costs of the devices.

[tinus42]

Hashrate wont continue to rise exponentially forever. At some point it will taper off and the share prices will go up again.

And what exactly is it that will stop hashrate from rising exponentially?

Hashrate increasing faster than Moore's law allows for?

[pand70]

Hashrate wont continue to rise exponentially forever. At some point it will taper off and the share prices will go up again.

And what exactly is it that will stop hashrate from rising exponentially?

Hashrate increasing faster than Moore's law allows for?

You do understand that Moore's law is just an observation right? For various reasons it works but it can't allow or disallow anything at all...
And hashrate has a long way to go since it will start get limited by Moore's law...

[Birdy]

You do understand that Moore's law is just an observation right? For various reasons it works but it can't allow or disallow anything at all...
And hashrate has a long way to go since it will start get limited by Moore's law...

It's an exponential increase, that's all you need to know to answer the question "Is it sustainable forever?" with no.

[moribana]

Hashrate wont continue to rise exponentially forever. At some point it will taper off and the share prices will go up again.

And what exactly is it that will stop hashrate from rising exponentially?

What stops it from going up exponentially forever is that in a finite system with finite resources you cannot get an arbitrarily high hash rate. Moreover, since the exponential is a very rapidly increasing function, resources will be exhausted pretty soon, so it cannot continue for very long.

But this does not mean that when the hash rate starts to stabilize anybody will still be profitable. In fact, there are indications that by the time this happens (in a year or so), everybody will be hashing just to limit their losses and nobody will make any profit. So I am not quite sure about the share price rising...