Bitcoin Forum

https://i.imgur.com/aFG3mBq.png

Last time it happened was in January 2013 after block reward halving.
That event more or less coincided with the start of an exponential rise in price.

I'm not saying "buy buy buy" just noticing the event has happened.

Sellers wising up? After all, why sell something at a loss?

I think AM turned off some devices to let difficulty cool down.

People just turning of their GPU's because the new ASICs take all the profit.
If you don't own an ASICminer, you will buy bitcoin

=> price goes up

Where is that image/data from?  Thanks.

http://bitcoincharts.com/bitcoin/

Why?!
Does not make sense.

Did some mining facility suffer some disaster and taken offline?

Yes, that doesnt make any sence, except somedy or a group of miners turned their asics off in agreemnt to ccol the hasrate down. Or maybe some big asic miners have issues with the sunny weather (acually about 35 C) in germany right now :-)

Do you have any ideas whats going on?

Here s another grapg from blockchaininfo:
https://blockchain.info/charts/hash-rate?showDataPoints=true&timespan=30days&daysAverageString=1&scale=0&address=

http://s22.postimg.org/6tumrr8sx/blockchaininfo_hashrate_drop_2013_06_19.jpg

---------------------

Asics turned on again. Everything back to normal :-)

http://blockchain.info/de/charts/hash-rate?timespan=30days&showDataPoints=true&daysAverageString=1&show_header=true&scale=0&address=

http://postimg.org/image/vwn9tsi1p/

It is only for two days I dont see the problem.  It may go up all over again

I think many GPU miners stopped mining since it's not profitable anymore with the current difficulty. Hashrate will go up again when more and more ASIC miners start hashing.

BFL stopped mining on their own and starts shipping? j/k

Maybe difficulty is dropping right after the last increase due to GPU miners giving up in higher electricity cost areas?  Right now most miners in Australia are priced out (US$0.24 to US$0.32 per kWh).  Some European miners would be struggling too.  As BFL continues to ship units we should see most GPU miners give up.  Only those left on some hydro and nuclear power will stay, or those with 'free' electricity.

Could be Avalon stop mining with customer machine too..

Maybe it's just that many GPU miners in the northern hemisphere have shut down (myself included) due to not being able to offset electricity costs by using them for winter heat.

No, GPU miners did not all turn off their rigs at the same time.

It was ASICMiner's hash rate that dropped. Probably swapping out its 2nd gen blades for 3rd gen. Or something.

ASICMiner shut off roughly 15 TH/s worth of mining equipment in the past 3 days. That is more than 10% of the world's hash power. That definitely had a big impact on it. I'm not sure why they did it by the way. It could be network or hardware problems, or maybe they don't want the difficulty to rise too much higher yet.

Anyways, it makes sense that we hit a plateau. ASICMiner blades are sold out. Avalon batch 2 still isn't shipping and BFL Jalapeno shipments have slowed compared to what they were 2 weeks ago. They also claim to have started shipping singles but no one seems to have one yet. In the meantime, with the last difficulty adjustment a few days ago many GPU miners suddenly became unprofitable and shut off their machines.


This plateau won't last though. Avalon should finish up shipping batch 2 in the fews weeks to a month or so. BFL looks like they will actually begin shipping singles finally, and ASICMiner will add more hashrate again. I think we won't see much change in difficulty at the next readjustment, but then it will gradually start to rise again, and we may go back to seeing 20%+ increases in difficulty a month from now. It all depends on when BFL and Avalon get their act together and ship their preorders.

The sky is falling!

There was a massive EMP set off by nefarious government agents designed to destroy the bitcoin network. Clearly they have failed.

Buy, Buy, Buy!

this is my bet. 

finally stopping the "burn ins".

^^^ THIS ^^^

seriously all of sudden a 500GHash unit shows up???

I'm going with these guys... BFL mined the bejesus out of customer machines and took it to the very brink of insanity/underhandedness. 

The bids already closed, but do you guys think this will have true or false outcome?

"On July 1st, 2013, Bitcoin mining difficulty will be higher than 20,000,000."
http://betsofbitco.in/item?id=1021

This right here.

Difficulty "recession" is knocking at the door.

http://dot-bit.org/tools/nextDifficulty.php

See last 120 blocks...

This rather interesting, but don't forget ASIC have only started to ship.. between avalon, bfl, and knc there are probably tens of thousands of units (ordered) waiting to ship...

Meanwhile, if you order from ASICMiner, you can have your hardware within a few days.

Methinks: so maybe 3 major ASICs producers have coincidentally stopped their pre-shipping "burnin tests", more or less around the same time.

KNC doesn't even have chips yet... Avalon has been recently exposed as mining with customer hardware with some credible evidence.  BFL claims they don't burn-in on bitcoin production net, but we all know what their word is worth.

 ;D +1 on that last sentence!

Still waiting for my single...  Should ship sometime...  Probably in one of the 'next week' that happen to fall within this year :D

If this is true (and I have no clue if it is) I wonder if they're trying to work the difficulty in order to help those purchasing their USB Block Erupters. There's been a lot of controversy as to whether they would ever earn an ROI at the prices they were selling at so I could totally see ASICMINER trying to lame-duck the difficulty in order to protect the interests of those who bought their product.

If this were the case, I would commend ASICMINER for looking out for their customers.

Purely hypothetical, just an interesting thought.

Losing out on $36000/day just to help a few people out? They'd have to really care about their customers for that.

True, there will be one retarget between now and July 1 and unless something very unexpected happens to the hashrate the new diff will be between 20 and 21m.

Could be just Northern Hemisphere summer and high temperatures. I wouldn't start drawing any economic conclusions here.

If difficulty drops low enough and price stays high, the GPU folks will turn their rigs back on.

Don't worry son...  Difficulty will rise...  As the ASIC start shipping :)  I think that whatever the BTC price does, at this point, just filling the pre-orders, the hashrate will rise.  And now, ASIC is a lot more profitable, power consumption vs return income.  So, even if price goes to 50$, or even 10$, ASIC will come!

If it were pre-ASIC era, then it would be another story.  Not a lot of folks are FPGA-mining.  GPU then would be the dominant miners.  But now, I don't think loosing even 50% of the GPU would make more then a dent on long term.  Well, in fact it is to be expected that GPU will phase out, unless BTC goes to 1000$ ::) or LTC and other scrypt coins raise relative to BTC (in which case we'll see GPU migration).  But until now, it's pretty staying near 1 to 1.

Diff. is taking a small nap before it boom right now.  Expect the next weeks to be 'fun' :)

You're right as to the fact ASIC's have only started, and no matter what, diff will go along. We're only seeing some caution now, where all companies have been accused (and proven) to excessively burn in their products, and they have now stopped doing it (or maybe still doing it but a much lighter rate..)

BUT, I think you're definitely wrong on the profitable assessment. In fact, if you take the current prices and delivery times of all ASIC suppliers, they only ones that will get ROI are the KNC ones due to the extreme difference in price; and only if (and it's a big if) they ship no more than 1 or 2 months late. Every single other device out there bought today is already unprofitable and will be so until ASIC makers cut their prices at least in half.

Not necessarily care about their customers. But imagine the future losses in sales from people burnt by their USB ASICs if they can't at least pay for themselves because the company that they purchased the hardware from is too big of a player. Not saying it would definitely balance out, but it might.

So true...  I feel sorry for peoples who still buy 50GH/s miners at twice the price you could get a 60GH/s miner a few months ago...  I would definitely wait for prices to drop before buying another ASIC now.  Being at the end of the line, when you receive your order, you won't make ROI.

The only one thing is that to still be "part of the game" in a few months, you'll need an ASIC.  All the rest will be like having a very power-hungry CPU today.  And I think it's safe to assume that once the difficulty goes up, and people calculate the ROI, then ASIC will just stop selling at the ridiculous prices they are now.  The prices will drop, and ASIC will get faster too.  Until then, as said, the high prices are sustained only because peoples are bad in math and basing their ROI based on today's difficulty, in an dream world where they would get their ASIC tomorrow, with current difficulty.

Aye. Also, think about the fact that ASICs are mostly (or entirely?) sold in usd, so if and while btc/usd goes down they become even more unprofitable at current prices.

So you got difficulty AND btc/usd both directly affecting ROI: diff quickly going up, and btc/usd trending down.

Looking bleak.

ASICMiner is the only one shipping with reasonable lead times.  They sell for BTC.  Anyone selling for USD is only selling you a preorder.

Yeah but they are not exempt from the crime. Take for example the usb miners they are selling right now. Approximately 1btc according to the groupbuy currently ongoing.

1btc = 330mh

That's absolutely ridiculous. That much hashing power with the current (somewhat slowed down) diff rate growth will never generate 1 btc, never ever.

Let's walk through this: http://www.bitcoinx.com/profit/

Let's do it in BTC terms, so set USD/BTC to 1 and price to 1.  Set electrical rate to 0.0009 (about $0.09/KWh).  Set hashrate to advertised 334.  Power Consumption: 2.5 W

Now, let's set profitability decline to 0.05, which is an assumption that we will have a 20x increase in difficulty over the next year.

Hardware break even   1 year, 1 day

Oh, and the ones I have in hand are hashing at 400MH/s, but shhhh.

https://docs.google.com/spreadsheet/ccc?key=0AmXn4BnLUgYsdEF0TUtWUWtUaUUzR1F2aUhEWW9lN2c#gid=0
at 10% every 11days
@ 1btc cost
you break even in a year. -2% for pool fee
keep in mind, last 12 cycles, average 17% increase

10% increase every 11days is optimistic.

http://www.coinish.com/calc/#

This is what I use and consider to be a lot more accurate in terms of predicting diff change.

It uses a formula based on the last 10 changes, so that gives you the lowest possible value of diff growth from today onwards, instead of a random 20x guess (which I think too low).

Your calculator is also not telling you that after that 1 year 1 day you won't generate more than another 0.1-0.2 tops for the rest of eternity. At most with calculator, you will break even.

Using my calculator you won't get ROI even if you were to use lower diff growth rate than the current one !!

Btw, if you already have the units working at home since weeks ago, that's a whole different story.

It is flat out wrong.

We're currently at 14,36% every 11 days.
 
And once ASIC start shipping by the thousands it will go up up up.. but let's say 18% to be optimistic..

Can you run it with both of those numbers?

Lol, you're being a dumbass.

Read the data. I'm not extrapolating anything.

Using the last 10 diff changes, the daily diff increase is right now of 1.3 %. These are current factual real values !

1.3 x 11 days = 14,36 %

Now go to that spreadsheet, fill in 14,36 instead of the (impossible)10%.

Voila. Unprofitable.

And you would have to be also really stupid to assume diff growth will be reduced by the entry of ASICs... but hey.. w/e go buy some more. I got ASICMiner stock. I profit off people like you.

You clearly don't know what extrapolation is.

Extrapolating is exactly what you are doing.

And you would have to ignore a lot to come to the conclusion that one company shipping reliably leads to an explosion in hash power while price is falling.  And don't forget they rely on foundries for chips that service customers like Intel, AMD, and Apple.  Good luck getting them to rush your order.

See how dumb you are:

original observation interval =


I used the original observation. No estimations here. Try it yourself. Unprofitable.

Now if remember correctly, your calculations are based on: the random assumption on 20x growth in a year. That is an estimate (actually maybe a guess at most..).

But, just for kicks, same numbers as before, but instead of 334 (advertised hashrate) and 0.05 (20X increase over a year), I will use 400 (actual hashrate) and 0.025 (a 40X increase).

Hardware break even   1 year, 5 days

Wow.  Where did you get your math training at?  You should ask for a refund.

You are claiming that the past directly predicts the future with a linear relationship.  This will not hold for the reasons I mentioned.

No dumbass. I'm claiming it will (can) not go down below the current level.

Imho it will go up by a lot, but for the sake of optimism I'm assuming it stays the same as today, for a whole year !

If you wanna do your calculations based on the assumption diff growth will reduce, by all means, go ahead.

I see why we are talking past each other.  You are talking about growth rate and I was discussing difficulty level.

I absolutely do want to base my assumptions on difficulty growth not continuing at it's current pace for a full year.  We will see it continue to rise in the short term, but without bitcoin price increases it will taper off.  Yes, if price rises enough to create a new difficulty boom you could indeed lose in bitcoin terms.  Barring that, there should be a gain over just holding tight, IMO.

You are entitled to your opinion, but I would prefer if you stuck to facts and logic instead of ad hominem.

Facts and logic were thrown your way a plenty, you misread them.

Sorry for going ad hominem on you tho.

Yes I did.  When I get bullied I get defensive and my thoughts aren't as clear.  Now please answer me this:  Who is going to buy all that hardware that is bordering on profitability at current levels?  Certainly people will get stop buying if difficulty continues to rise at the rate you claim.  Won't that slow down growth?

Well, a few things can happen, but basically ASIC companies have to adapt and respond. They are big market players now, with lots of power.

If people stop buying, ASIC companies have to lower prices (they operate at huge margins when you consider the "burn in's"). Mining profitability increased, people will buy again.

If ASIC companies don't wanna lower their prices, they can put them to mine themselves, we all know they've done it already.. It would be a total mess though..Idk what would happen to btc, prolly not good.

There are less than 1000 TH combined between all preorders and inventory (in the case of ASICMiner).  Currently the network is hashing at 150 TH/s.  Even if all of it came online, we would be less than 8X current difficulty.  Is there really enough money left to be invested into ASICs to push us past 20X in the next year?  That's another 2000 TH of ASICs that aren't even planned for production yet.

1000 TH is way off.

KNCMiner alone has already more than 2600 orders (last time I checked, maybe a week ago, bound to be more now).

The big model has 350 GH/s and the small one 175 GH/s . Lets use that one:

2600 * 0.175TH/s = 455 TH/s

If I do a wild guess and say they sold half and half:

1300 * 0.175 TH/s =  227.5 TH/s
1300 * 0.350 TH/s =  455 TH/s
TOTAL = 682.5 TH/s

That's KNC alone, and they are the newest player, and the ones who've been taking orders for the least time.

Okay
KNC 700
BFL 800 (this (http://bfl.ptz.ro/) lists 408 and some are already deployed, but lets double it for kicks)
AM 200
Avalon 300 (pretty clueless here so I based it on about ~3TH per 10k chip order X 100 orders)

Even if we assume everyone delivers within the next year (Avalon has been iffy lately, and KNC has yet to prove themselves), that is 2000 TH, 1500 of which is already purchased waiting for delivery.  So, we can absorb 500 TH/s of demand before anyone needs to produce another run.  Is there really another 1500 TH worth of demand for miners bordering on profitability?

I get what your saying and if those are the real numbers or even a little below, things may go just a tad smoother.

But still, think about the following. It's not really about any of the companies actually needing to produce another run.

This is a game of profit, for everyone.

As long as there a sliver of profit in mining gear (or even when there is not..), people will buy them.

If buying stalls, ASICs will still want to keep raking profits, so they either lower the price or iterate or both: New Model ! 2x Power, Half Energy Cost, Half Price ! (obviously exaggerating here.)
There's always profit in maximizing efficiency, so there will be demand for that; and remember, we're in version 0.01 of the ASIC era...still lots to learn and improve...

I'd say sooner than later there will definitely be a correction in ASIC prices. These guys profit on sales, not on a stalled market.

That'll keep things moving, and that's really the end scenario imho. As long as there is profit to be made from any party, the market will not stall.

 

To modify the mantra of Steve Balmer: "Delivery! delivery! delivery!" Promises of GHs being delivered to the network have proven in the past to be utterly unreliable. You cannot quote KNCs promise to deliver as a fact before it happens. Assigning some sort of probability to their hash-rate to the network would be prudent.

I also often worry that by selling at unprofitable prices the ASIC companies are creating a huge time bomb of angry small and big time investors new to btc who bought in on the promises.

I have no doubt in my mind, there will be huge backlashes in 6 months.

Heck, there already are if you consider BFL; but that's just a bad company I guess..