Topic: Hash rate declining slightly in recent weeks. (Read 4372 times)

3  blocks to go, and it looks like the next difficulty adjustment will be about +1%.

Yup, heh wake me up when the difficulty gets to 50,000,000,000 or a network hashrate of about 400 PH which ever comes first...

fixed

You might want to fix the bitcoinwisdom link.  

back from helping to nurse my sick brother-in-law back to health.

Real world stuff sucks much nicer typing away on site.

and  on :

https://bitcoinwisdom.com/bitcoin/difficulty


Bitcoin Difficulty:   34,661,425,924
Estimated Next Difficulty:   35,474,047,220 (+2.34%)
Adjust time:   After 246 Blocks, About 1.6 days
Hashrate(?):   271,806,654 GH/s



and on http://bitcoincharts.com/

Difficulty   34661425924
Estimated   35000997201 in 246 blks

about 0.97%

What? Are you talking about recent times? There has been no drop that has lasted several weeks since the introduction of ASICs. If you're talking about this difficulty period, there has been no drop in hashrate because the difficulty is going to rise in a day... unless of course the whole network has been statistically lucky during this period. But then you'd have no way of knowing if there's some phantom datacenter that has shutdown or whether it was just luck.

People are reading too much into daily variations.  The network hash rate calculated over 24 hours is only an approximation. As a rough rule of thumb assume up  to a 10% error over 24 hours.

For example if the real miner network hash rate is exactly 250 petahash, it may show as 225 petahash one day, and 275 petahash the next.  Nothing has changed.  It is just statistical variation.

Over 2016 blocks (one difficulty change) the error is less.   Each difficulty change will have an error up to say 3%.  So we will soon start getting some negative difficulty changes as the network growth slows, even if just due to statistically variation.

If you want the mathematically details here they are. There is a 10 minute block average solve time with a 10 minute variance (poisson distribution).  So the standard deviation percentage error in solve time is 100 / sqrt(number blocks).  This gives a standard deviation error of 8.3% for 144 blocks (1 day), 4.5% for 504 blocks, and 2.2% for 2016 blocks (difficulty change).  The variation will be within 1 standard deviation of the true value 68% of the time and within 2 Standard Deviations of the true value 95% of the time.  
  

There's been a pretty steep drop today. Coincidence, are people trying to influence the imminent adjustment, uncompetitive miners shutting down?

There was a drop that lasted for several weeks. That's not just noise from the rate at which blocks are found.

It's possible that some big hashing center was doing an upgrade, and they were shut down during the upgrade.

There is no magic network hashrate chart that shows exactly how much is hashing on the network at any given time.

The network hashrate is determined by how quickly blocks are found within the current difficulty. If a block is found every 10 minutes (6 per hour) on average, the chart will show no growth. But if there is a steak of good luck and 12 blocks are found in an hour, the network hashrate will have appeared to have doubled even if it hasn't.

That's an extreme example but it illustrates a point. You can't look at the chart for a short period and tell for certain whether it's actual growth or just luck. But you can make an assumption that if you are seeing short bursts in growth that quickly disappear, it's most likely just luck. No massive mining operations are randomly turning 50PHs off and on. It makes absolutely no sense not to just continually mine. Nothing is gained by shutting a whole datacenter (or two) down.

We are speaking of hash rate not luckiness, wth!
What i saw was over 50ph added in an interval of ~3hours. That's what blockchain.info showed on their site.

Maybe it was more than 1 farm, something orchestrated...
Maybe it had something to do with the that whale selling 30k at 300usd...

Bitfury and Bitmain aren't "testing" anything. They're just building massive datacenters.

like others said it is most likely big companies like bitfury, bitmain, etc. Testing products before shipping etc. They control thr hashrate since most miners do not have that type of hashrate to see thr fluctuation. if some home miners turn their miners off it would not even scratch the difficulty to make a diff

Agreed.

Oh for god's sake, quit spamming you betting site.

This! Do not underestimate variance. The blue line at: https://bitcoinwisdom.com/bitcoin/difficulty has always gone up and down. In VERY REAR cases this was because the real hashrate has dropped. Remember that no one knows the real hashrate of the network. It is always just an estimate of the current frequency at which blocks are found. And this is after all a random event.

We try to provide the best numbers for future difficulty changes here:
https://www.fairlay.com/event/category/bitcoin/difficulty/

The likelihood of a decrease is very close to 0.

that is most likely part good luck variance.  but some believe   that bitfury can add 150th at will or just a flick of  a switch.  my guess is the 246 was a little low due to bad luck.  the 308 is high due to good luck.  the real number is 260 or so.

think of it as tossing a pair of dice  throw 3 sets of box cars in a row (6 + 6)   hard to do but possible.   so a swing from 246 to 306 is part that.

growth looks to be under 5%  when this settles

All this new hardware going online and btc price keeps going down.

Panic sellers and miners of all sizes needing to sell to cover costs, things aren't looking good

It's much more than that:
https://blockchain.info/charts/hash-rate?timespan=30days&showDataPoints=false&daysAverageString=1&show_header=true&scale=0&address=

246ph to 308ph!!!
few hours ago it was ~246ph

And the prismas didn't left china yet
So it's a very big farm, or a few smaller farms owned by same group/person!

 

no one, what it is:

the s-4 shipments are arriving to buyers and being put on line. >>>>>>>>>>>>>I found at least 600th in btc orders. 


the asicminer long tubes are arriving to buyers and being put online. I found at least 100th in btc orders. 

  this is in the open but many other sales are made with cash and are hidden to me.

more s-4's  above are going to hashnest.

and asicminer may be mining quite a few longtubes privately

Suddenly, in the last two days, the hash rate went way up. Somebody just brought substantial new mining capacity on line. Looks like the difficulty will not go down.

Any idea who has the new big mining farm?

The chart in your link is now showing +1% difficulty change. So much for the "difficulty is going down" crowd. And still 5 or 6 days to go. 

https://bitcoinwisdom.com/bitcoin/difficulty

Big farms? The "big farms" are the ASIC manufacturers. Their equipment costs are a fraction of what retail machines cost. And their equipment is hosted in areas of the world where they pay a fraction of what your average home miner pays for electricity.

IF the price continues to slide, I think you will see the big farms start to sell off their equipment as they need the price to be at a certain point to be profitable. When they purchased the equipment, the btc price was probably above 600, so a 300 dollar price is pretty scary for them..

A difficulty decrease would be awesome but I'll believe it when I see it. I turned off all My S1s a while back. They were costing more in electricity than they were making in Bitcoin. Would make me very happy if I could turn them on again.

Where did you get the chart?

We need to drop the BTC/USD to 250, so 30% of the 'cloud' miners stop. Then it will be nice.

hash rate going up again, so yeah, most probably will be a small increase.

I think we will see a small increase in difficulty this next time.

We're halfway to the next difficulty adjustment, and the green line (difficulty if it was adjusted right now) is still below the red line (current difficulty). (The grey line (recent hash rate) is climbing again, though.) So we're on track for a decline in difficulty, or at most a small increase.

I believe its due to BFL...

Punishing and greed is the wrong frame IMO. Home miners are no less greedy than industrial miners, and profit motive isn't a bad thing. Home miners are simply having their asses handed to them, and uncompetitive data centers will also go out of business. This is simply a case of market forces in action, and it's healthy. I really doubt home miners will be able to maintain a significant share of the total network hash rate and unfortunately this may lead to more centralisation. Mining has become a business, and the rational thing for miners to do is to sell off all their bitcoins. Companies can also speculate in BTC, but these are really two independent business processes. There will be no return to the golden age of accumulating coins cheaply through home mining, just as there will be no return to GPU or CPU mining, no matter how much angry people with financial decisions that didn't pan out complain about it. There's no such thing as an indefinite free lunch, if you want to earn money you have to do something for it: perform competitively priced labour, bear financial risk, or save some money.

Agree. I think home miners who are marginally profitable should hang on to their equipment. If you are paying $0.22 for power, then maybe not so much. If you believe in BTC long term, then lower prices are good short term.

The fact that hash rate is no longer increasing even though we know at least some brand new equipment is coming on line means we start to understand the economics of the older mining farms. Some of them can no longer pay the bills at BTC less than $400. By pay the bills I meet are cash poor or choose not to operate equipment at a loss. So some of the 2013 equipment is getting turned off. At a power plus marginal operating cost of $0.12 GHS, a 2W/GHS miner is no longer profitable. If you saw my post about Chinese power prices, they are all over the map. If anybody knows when the power cost of hashing dropped below 2W/GHS, please share it with the group.

IMHO, this means that at least some large mining operations have to regularly sell their BTC to keep the lights on. Mine and hold is not an option to them. As discussed in other posts - this becomes a death spiral until miners start shutting down-thus stabilizing the drop in BTC price. I see at least two possible dynamics:

1. A farm is still at break even on the fixed costs with newer equipment. It is just the older equipment operating at a loss on marginal operating costs. This means the marginal cost of operating the older equipment is just the power (and a bit of labor). That equipment can be turned back on if difficulty drops or BTC price increases.
2. A farm does not have enough of the right type of hashing power to cover fixed and marginal costs and are shutting down either voluntarily, by the power company or by their landlords. These farms are much less likely to come back up if operating profitability of BTC mining improves.

In either event, as I have mentioned many times, the current economics makes investment in a new generation of ASICS highly problematic for the mining companies. I do have an MBA to go with my Physics degree. The business case would not pass muster here in the US. If I am right, then we may see a fairly steep  drop in the rate of difficulty increase as long as BTC price stays low.

Further, there is nothing to put upward pressure on BTC prices because industrial miners now dominate production. They have to sell.

As the OP said bitcoin ecosystem is punishing greedy mining farm companies. Delicious.

God I hope so    Come on $100 BTC!!!!!!!!!!!!!!!!!!!!!! 

more mining--->more sell pressure on btc---->>more difficulty--->more electricity cost----->no Return of Invest----> and shut down with negative profit ----> sell miners HW under pruduction's (or buy) cost ---> underinvested miners win and worst ( assume this word widely ) HW producers and big farms lost

cicle will repeat when people will (and they will) buy equipement for double (or more) the production's cost, probably because the next price's bubble

the cicle won't repeat when speculators understand that high ROI in bitcoin world is allowed only for very shot timeframe and at a huge risk.

However I don't think the time for true difficulty drop is now 

that big mining farms,companies had just punished themselves..

more mining--->more sell pressure on btc---->>more difficulty--->more electricity cost----->no Return of Invest----> and shut down with negative profit

bitcoin ecosystem is punishing greedy mining farm companies..

I'm sure hashrate is declining, with the cost of electricity in most countries and current difficulty / btc price its no wonder.

Mining is not profitable in most countries anymore even with the latest hardware.

I say not profitable. Most miners are not breaking even or in other words, they're losing money.

Price drop plus bfl bust plus variance equals first diff decline is years. Let's hope

i'm sure its the bigger coporations giving in now that the price is low. but as soon as it goes up so will the hashrate

Ya something had to give.

This is geart news, for me

Hash rate, recent weeks.

The cost numbers have been indicating that it's time for some miners to drop out, and that seems to be happening.

If the green line stays below the red line, the difficulty will drop at the next adjustment.