Bitcoin Forum

First block of current difficulty: http://blockexplorer.com/block/00000000000006f5d58e11e705c01b33a87f6d31f6b5bded34f01c68ad4e3d61

Initial estimates of next difficulty are, of course, crazy, we need to wait a while for them to stabilize.

who said it was going down?

The people who were saying that difficulty follows price.

Some were speculating it could go down.

The estimated difficulty was tanking over the last few days, presumably because of miners pulling their hardware and selling it, and I'm sure the larger pools getting DDoSed didn't help. I'm guessing some people extrapolated that and combined it with their own desires to thinkthat the difficulty ought to drop, but they failed to take into account that in the first half of this recalculation period, we had huge growth of hashing power.

I'm not sure if we've been in the new period long enough, but BitcoinCharts is forecasting a growth in difficulty of only ~2.4k right now.

It hasn't caught up to the price yet.

We shouldn't ignore that the percent difference between this difficulty and the last has decreased by over half in comparison to the previous percent difference. the difficulty is definitely going to plateau in the coming weeks.

^^ this.

Right now next dif estimate is at about another 10% increase, or ~1.727mil  
This does not account for the large amount of hardware still being bought that is not online yet and the lag behind price which has recouped some today.



edited; for crap typing skillz

Makes sense, yet at the same time it's an arguable subject.  Who's to say that price doesn't actually follow difficulty?  But the price just "hasn't caught up yet" either.  Chicken or the egg, you pick.  :)

Correlation != causation, but the sipa charts show growth trailing off or even possibly going negative:

http://bitcoin.sipa.be/

I think it rather follows value. Value is not always the same as price. For instance, as I think Bitcoins are going to rise by several factors in the next years and I think long term their value for me is much higher than it's current price.


Difficulty affects price and vice versa, but there are other factors to consider as well.

Usually when I see something be referred to as a Chicken or the egg problem it's simply a misunderstanding of logic or the nature of the specific underlying problem. This is no exception, what you're talking about is a feedback loop, there is nothing weird about it and is omnipresent in everyday systems, including our body. The "Chicken or the egg problem" itself arises from inconsistent definitions, it's the same as asking whether a fixed object would move when pushed by an irresistible force.

Revised 2011-07-06 21:13 UTC: formatting & wording.

Well, difficulty rose even before Bitcoins had a $ amount affixed to them.  So literally speaking, difficulty came first, followed by price.  Somewhere along the line though, you're right that the argument became more of a feedback loop.  And that's probably a better analogy to use at this point in the game.

Definetly agree but not nearly as fun.  I vote chicken. Wait no, I vote egg..  Does a partially devoloped embryo count as both? ;p

The network did shrink a little for the first time. But because difficulty is often delayed (it averaged the entire last two weeks, which it not where the network currently is), the difficulty didn't go down. It might go down next time if there isn't a big rally between now and then, but I'd give 3:1 odds that there will be.

http://bitcoin.sipa.be/growth-small-10k.png

Notice it poking under 0 recently in the 3-day growth rate. Although, that might have something to do with the recent DDOS'es. In any case, I'm happy, good mining.

A good portion of the down poking of the past few days can likely be attributed to the Ddosing of some of the larger pools and the ~60Ghash that switched to mining namecoins for a few days. And a small percentage of those 'at-home' types flipping the miners off.  Albeit that last one is likely very small in light of the whole, as is the 60Ghash estimate, it all adds up.

I'd take a good look at it again in another 7-10 days, giving time for new recently acquired hardware to come online and other factors(price increase again) etc.

My vote is it going up enough to drive the next difficulty estimate up to about 1.8mil by then from the current 1.727 estimate.

At the current ~240 % profitability (typical profit/cost of electricity), I wouldn't expect any long term decreases anytime soon.

At 1 BTC = ~$15, I'd expect it to increase up to roughly 5,000,000 before plateauing.

You have to factor in equipment and space too. Electricity isn't the only cost to mining. I would think it lower, 3.5 mil, perhaps. And most big mining operation are counting on the continued success of bitcoin to justify initial costs. I think the past months troubles have made some question that.

And it's the summer holidays for school children now - so more tiem taking over home computers to play video games. Yet another small factor? :)

5,000,000 difficulty @ $15 per Bitcoin would give you a little over 10 % markup, which is still somewhat above where most highly competitive markets settle. Initial costs are only a barrier to entry (tying up capital), and operating costs are typically tackled with economies of scale. Obviously, at less than $.30 profit per day per gigahash, you'd need a giant operation for mining to make any sense, anyway.

The numbers are very rough estimates, of course. For instance, I'd expect serious mining operations to by power in bulk or do something more creative such as start their own windfarms in order to get some cheap electricity...

You also have to factor in hardware depreciation costs.. New video cards come out all the time, usually once a year but can be sooner... and also the fact that other technology could come out and disrupt your entire "large scale" operation (ASICs etc). Technology moves very fast - it's going to have to be higher than $.30 per gigahash. At that rate you'll almost certainly go negative, unless you play the markets and make sure to sell your hardware at the right time.

It's a mistake to think in terms of USD profit. Did you forgot than Bitcoins are themselves valuable?. Furthermore, even if you don't care for Bitcoins, no one is obligated to sell at the current market prices, that's actually the stupidest way to go.

I think only means it will get more difficult to even enter the mining business without the capital to put into the hardware. Which is really no different from any other type of competitive business. Only the businesses which invest in efficiency survive.

The profit is easiest measured in the currency that you pay your electric bill in. But since they are exchangable, it doesn't really matter.


So who is buying above current market prices?

At least for the time being, I think most people/companies who approach mining as business need to exchange the generated coins to another currency in order to pay the bills.

Why? Chances are good (if you're in the USA) you bought your equipment in USD, and you almost certainly pay your utilities in USD. It also makes a good frame of reference, even if you do only consider the things you can buy in BTC.

If Bitcoins are worth $15USD, most people who are selling stuff are going to price an item that they value around $15 at or near 1BTC. So I don't really see how not thinking about USD helps at all, if anything it just throws off your frame of reference considering the wildly fluctuating value.

1BTC right now is not the same as it was a while ago, and it's probably not the same as it will be in a month (for better or for worse), the changes in value of things measured in USD changes much less dramatically.

He is talking of mining and holding, hoping for the price to go up. That is a valid strategy, but if you are counting on the price going up, you can often make more if you _buy_ and hold instead of _mine_ and hold. With rising difficulty, many cards will not ever make as many bitcoins as you could have bought with equivalent funds at the time. This is particularly true is you time your buy well, like the $11 dip yesterday.

The price isn't guaranteed to go up again any time soon, IMO. Limiting supply doesn't help if there is no demand. The market would simply halt, if everyone just held their coins in hopes to get the price up.

In fact, in the medium term, I'm pretty sure there is a large amount of Bitcoins just waiting to be sold, which is likely to keep the price down for quite some time. This is actually good news to Bitcoin but bad news to those who thought it was a way to get rich quick with a bunch of Radeon cards.

More likely, if it reaches $0.30 per day per ghash, it will be more like F@H. Only enthusiasts will support the network. The rest will just buy. Of course, this means less security on the network overall.

Some of the best efficiency ratios I've seen is around 2.2Mhash/Dollar. That's still around $500 per ghash. 1 Ghash would be $108 per year. You're still negative by $392, after a year.

So you'd have to sell the hardware for more than $400 to make a profit after 1 year of use. Not likely in the tech world.

I said profit... Meaning something like $2.70 operating costs + $0.30 profit (roughly 10 %) = $3 per gigahash per day. Obviously these numbers are just very rough guesses but shouldn't be too far from the reality.

High difficulty and low security can't coexist AFAIK....

Maybe I'm using the terms wrong, English is not my native language.

If it were at $0.30 per day per ghash, that means that either the prices is really low, or the network is HUGE. If the price is low, then the system didn't take off, security isn't a big deal. If the network is huge, then it is secure.

Well, the main thing is..

Higher profit per 1Ghash = more miners. (higher security)

Lower profit per 1Ghash = less miners. (lower security)

Every miner has a certain profitability they're trying to reach, otherwise they will drop out.
Sort of seeing this already, even at the current profitability. Network growth has basically stalled.

In the future, I can sort of see ultra high efficiencies in mining, with huge companies supporting the entire network, but that's years away. Bitcoin would have had to reach some sort of ultra-stability and high trust for that to happen.

No, I don't see it that way.

More miners = higher difficulty = lower profit
Less miners = lower difficulty = higher profit

The network is self balancing. There will always be a number of miner sensible for the amount of money in the system.

Anytime there's a boost in profitability more miners will show up. That's just the way it is. Some will stay others will leave after a difficulty increase, but overall the more profit there is the more miners there will be.

If there is less miners, that means the profitability crossed their threshold and they left. As soon as more profit shows up, there will be more miners again.

But there are more miners now than there have even been, but profitability is quite low compared to the past.

And if miners show up when it is profitable, it will just increase the difficulty, decreasing profitability.

Anyway, we are just arguing in a loop. The system is well regulated, when there is a need for more miners, it will be incentivized.

Right, like I said, every time you go through an area where there's HUGE profit (like when bitcoin was at $30) people will buy tons of hardware and go through all the setup and everything. Once they get up and running, as long as there's reasonable profit for them and doesn't cross their threshold, they're not just going to pull the plug offline right away. So the profit does shrink every time this happens, but the network grows in security.

I believe that bitcoin mining difficulty lags price, because dramatically higher prices strongly motivate miners to increase capacity, which leads to higher difficulty.  Likewise, I predict that if prices decline for a while, then mining difficulty will level off or perhaps decline as some miners lose interest, or conclude that new capacity will not have a prompt pay-back.

chodpaba has actually estimated a difficulty decrease two re-targets from now

http://forum.bitcoin.org/index.php?topic=13339.60

He could be right, things have been slow on the prices-crazily-moving-upward front.

More miners decreases the profit per miner. It's actually about the ratio between difficulty and value of BTC, and where it will eventually settle.

I argue that mining profitability will eventually match others similar markets, which means that at the current BTC valuation, difficulty would be somewhere around 5,000,000.

Currently, you are getting back more than twice what you pay in electricity, which should still be very lucrative to many. But you need a lot of hashing power to make meaningful money.


I think that has more to do with the DDoSing of the pools than miners actually dropping out.