Asic difficulty newbie question.

[Myrddindc]

Hi all,

This may be a silly newbie question, but I'm going to ask anyway!

Everyone seems to be implying that Asic chips will increase the difficulty (I've read between 5x current and 20x)

But it seems to me that if difficulty goes up, those with inefficient mining systems (cpu/gpu and possibly some fpga's)  will be minning at a loss when factoring in power consumption. If this is true wouldnt they simply quit mining? Thereby lowering the difficulty. . . .

I often read on this forum that the current difficulty makes mining only barely profitable (or are people just trying to stop newbies from going into mining?) if so, it would only take a very slight increase in difficulty (say 2x) to make exisiting mining completely unprofitable . .  And so overall difficulty will not increase by as much as everyone seems to expect?

I realise this is based on a lot of assumptions. But I would apreciate a reasoned logical response as I am only curious!

Thanks!

Myrddin.

[DeathAndTaxes]

They will quit however total has rate will still go up.  As an example if 300 TH/s of asics replace 25 TH/s of GPUs then difficulty is going to increase by more than a factor of 10x. In the long run once multiple vendors start shipping in volume difficulty is going to go up a lot more.  50x current is certainly possible.  Given the break even time for a new GPU rig is 6 to 12 months trying to jump in now with obsolete tech is a bad idea

[Myrddindc]

Ahhh, I see, so the total hash rate of the new equipment, even if in the hands of only a few will be about 20 to 40 times the current, and will dwarf the current equipment.

Reasonable difficulty estimate in 6 months on this 10x current?

Thanks again!

Myrddin

[nathanghart]

At 10x difficulty I can't create a scenario where mining is even close to profitable to anyone that is not getting their miners at cost. Will we see a dip in asic miner prices or how does the community survive?

[DeathAndTaxes]

At 10x difficulty I can't create a scenario where mining is even close to profitable to anyone that is not getting their miners at cost. Will we see a dip in asic miner prices or how does the community survive?

"Profitable" or "my unrealistic expectations is to break even in 2 weeks and make 487947298721472987892472% profit per year"?

Capital Cost: $1500 + $500 (unit plus one module upgrade to improve efficiency)
Electrical cost (1 year): 800W * 24 * 365 /1000 * $0.08 = $560
Total lifecycle cost (assumming 12 month life): $2,560.64

88GH/s @ difficulty 30M (~10x current) = 538.2 BTC per year.

$2,560.64 / 538.2 BTC = $4.75 per BTC.  If you could pay me $2,560 right now and I would pay give you 538.2 BTC over the next year would you?  Of course you would.  If you think difficulty will stop at 10x current you are basically saying once it gets there people will say "hmm only 400% profit per year, nope I don't think so".  Of course difficulty will go way way beyond that.

Even at difficulty 100M (~33x current), 88GH/s = 161.4 BTC per year
$2,560.64 / 161.4 BTC = $15.80 per BTC.  I still think thousands of people will take that deal.

Note both those numbers assume difficulty jumps instantly.  In reality since difficulty will rise over time the profitability is even higher.  Granted the ASIC might not be cost effective for a year so there is the risk element but the risk has never stopped miners before.   Anyone thinking difficulty is only going to rise 2x or 5x is simply dellusional.  Even "only 10x" is likely way too optimistic. Difficulty 100M is likely the min realistic value and if BFL power efficiency is true and they can ship in volume ... 200M or higher is a very real possibility.

[nathanghart]

At 10x difficulty I can't create a scenario where mining is even close to profitable to anyone that is not getting their miners at cost. Will we see a dip in asic miner prices or how does the community survive?

"Profitable" or "my unrealistic expectations is to break even in 2 weeks and make 487947298721472987892472% profit per year"?

Capital Cost: $1500 + $500 (unit plus one module upgrade to improve efficiency)
Electrical cost (1 year): 800W * 24 * 365 /1000 * $0.08 = $560
Total lifecycle cost (assumming 12 month life): $2,560.64

88GH/s @ difficulty 30M (~10x current) = 538.2 BTC per year.

$2,560.64 / 538.2 BTC = $4.75 per BTC.  If you could pay me $2,560 right now and I would pay give you 538.2 BTC over the next year would you?  Of course you would.  If you think difficulty will stop at 10x current you are basically saying once it gets there people will say "hmm only 400% profit per year, nope I don't think so".  Of course difficulty will go way way beyond that.

Even at difficulty 100M (~33x current), 88GH/s = 161.4 BTC per year
$2,560.64 / 161.4 BTC = $15.80 per BTC.  I still think thousands of people will take that deal.

Note both those numbers assume difficulty jumps instantly.  In reality since difficulty will rise over time the profitability is even higher.  Granted the ASIC might not be cost effective for a year so there is the risk element but the risk has never stopped miners before.   Anyone thinking difficulty is only going to rise 2x or 5x is simply dellusional.  Even "only 10x" is likely way too optimistic. Difficulty 100M is likely the min realistic value and if BFL power efficiency is true and they can ship in volume ... 200M or higher is a very real possibility.

You are right of course, I did not take the time to do the math and posted in a hurry from my phone. It looks more like 40x increase over the next year (based on certain estimates almost no one can verify with certainty) which would still be profitable, but requires a large investment. I guess my question was more about the future state of bitcoin. If only asics are profitable does that somehow reduce the interest in bitcoin unless asic prices come down. Will the adoption of bitcoin go down since not everyone can get into it with existing hardware, minimal investment? Where do we see the growth in terms of people coming from D&T? I suppose it has to be a new disruption in a current market, be it online gambling, mmo currency, donations, or something else? What do you think?

[DeathAndTaxes]

I don't get where this idea that ASICs means hundred thousand dollar hashing farms?

Avalon makes a $1500 unit.  BFL has 3 different sized ASICs under $1500.  Try building a multi GPU rig for under $1500. Obviously ROIC is based on the capital invested but I don't see how/why a $150,000 ASIC investment is going to have a higher % return than a $1,500 or even $150 one.  Honestly today you aren't going to make more than token few bitcents with anything other than a high end dedicated multi-GPU rig.  My rigs (offline now) are quad 5970, 1250W custom built rigs.  The amount of time, and capital goes way beyond a $1,500 investment.  

Still mining is old and becoming a commodity business.  All hashes are the same, there is no brand differentiation or barriers to entry.  The margins are going to get smaller and smaller and smaller.  It has been done and perfected.  The whole rest of the economy is an almost empty canvas.  All it takes is a little creativity.  Think the inventor of Satoshi's Dice wishes he had spent more time mining a few low margin bitcents out instead of wasting his time building that site which generated a half million in profit in six months?