Mining difficulty slowing down over the next 3 months?

[DolanDuck]

A  Big chunk of the profits being mined are re-invested into more equipment to offset the rising hashrate.  But the initial investment is not 100% recouped.

Mining bitcoins at home, for people who are capable only of low investments, is almost ended because of the small or null ROI they can have.

[s1gs3gv]

s1gs3gv's law:

"As long as bitcoin represents a useful form of wealth storage for the wealthy the price of bitcoin will adjust such that decentralized mining remains profitable."

[taipo]

The next leap looks to be a big-ish one...

[ujka]

The next leap looks to be a big-ish one...

Bigger than the last, most likely yes. But not biggish enough to be inbolded on https://bitcoinwisdom.com/bitcoin/difficulty

[ElGrandJefe]

The network peaked at nearly 100 PH just before the last rise in difficulty, then plummeted, and is now just over 82 PH. Unless it comes back, or a lot of other hardware comes online, this could be one of the smallest difficulty increases in recent memory - or even a slight decline.

Why did nearly 20% of the network go offline? Any speculation?

[ujka]

No hardware went offline. It was just 'luck' in generating new blocks. I saw payouts from ghash pool about 30% over average on that spike.

[davejh]

The network peaked at nearly 100 PH just before the last rise in difficulty, then plummeted, and is now just over 82 PH. Unless it comes back, or a lot of other hardware comes online, this could be one of the smallest difficulty increases in recent memory - or even a slight decline.

Why did nearly 20% of the network go offline? Any speculation?

It probably didn't (although there's no way to be absolutely certain) - the statistics of mining mean that short term huge positive and negative spikes can and do occur. There was a much bigger one a few weeks ago that had me go and look at this in a little more detail. Even over a 2 week period it's quite likely that we'll see +/- 5% changes in the difficulty even if the total hashing rate stayed the same.

I put the 2 writeups here:

http://hashingit.com/analysis/16-hash-rate-headaches
http://hashingit.com/analysis/17-reach-for-the-ear-defenders

[taipo]

It could be that manufacturers are doing burn in tests. Big mining operations are ordering to keep ahead of the diff. So it could be a timing thing, 20 PH is ordered, burn in test takes place, miners are shipped and arrive at their destination just before the next diff rise, meanwhile in anticipation of the next, large mining ops order more, more burn in tests, more peaks just before the change.

If you go through the Eligius contrib list, you can see this rising in hash power just before diff changes in most of the medium to larger mining operations.

[samsonn25]

It could be that manufacturers are doing burn in tests. Big mining operations are ordering to keep ahead of the diff. So it could be a timing thing, 20 PH is ordered, burn in test takes place, miners are shipped and arrive at their destination just before the next diff rise, meanwhile in anticipation of the next, large mining ops order more, more burn in tests, more peaks just before the change.

If you go through the Eligius contrib list, you can see this rising in hash power just before diff changes in most of the medium to larger mining operations.

20PH is alot!  I dont think any one manufacturer has that on hand right now to test.

[lightfoot]

It could be that manufacturers are doing burn in tests. Big mining operations are ordering to keep ahead of the diff. So it could be a timing thing, 20 PH is ordered, burn in test takes place, miners are shipped and arrive at their destination just before the next diff rise, meanwhile in anticipation of the next, large mining ops order more, more burn in tests, more peaks just before the change.

If you go through the Eligius contrib list, you can see this rising in hash power just before diff changes in most of the medium to larger mining operations.

20PH is alot!  I dont think any one manufacturer has that on hand right now to test.

Agreed. 20ph would require what, 20 megawatts of power? That's kind of a lot.

1gh=1kw. 1000gh=1ph=1mw.

[ElGrandJefe]

The network peaked at nearly 100 PH just before the last rise in difficulty, then plummeted, and is now just over 82 PH. Unless it comes back, or a lot of other hardware comes online, this could be one of the smallest difficulty increases in recent memory - or even a slight decline.

Why did nearly 20% of the network go offline? Any speculation?

It probably didn't (although there's no way to be absolutely certain) - the statistics of mining mean that short term huge positive and negative spikes can and do occur. There was a much bigger one a few weeks ago that had me go and look at this in a little more detail. Even over a 2 week period it's quite likely that we'll see +/- 5% changes in the difficulty even if the total hashing rate stayed the same.

I put the 2 writeups here:

http://hashingit.com/analysis/16-hash-rate-headaches
http://hashingit.com/analysis/17-reach-for-the-ear-defenders

Thanks for the links. I'm well aware of the "luck" component of mining, and I know there's a lot of statistical noise out there, but the nearly 20% drop really puzzled me.

There's no way to know the true cause, of course. Maybe Friedcat (or someone else) was testing a batch of new equipment before shipping it. Or maybe it was just a huge statistical deviation.

It particularly struck me because I've noticed that recently (meaning the last 10 or so difficulty increases) the reported network hash rate tends to peak/spike a day or two before the next difficulty increases, then drops off again - almost like some big players are trying to game the difficulty rate. The hash rate picks up again after the difficulty increase.

If that's really just statistics noise, there seems to be a fairly reliable pattern to it. But this time around the hash rate peaked around 2 days before the next difficulty adjustment and then kept dropping. It's back up to ~88 PH/s now.

[davejh]

Thanks for the links. I'm well aware of the "luck" component of mining, and I know there's a lot of statistical noise out there, but the nearly 20% drop really puzzled me.

There's no way to know the true cause, of course. Maybe Friedcat (or someone else) was testing a batch of new equipment before shipping it. Or maybe it was just a huge statistical deviation.

It particularly struck me because I've noticed that recently (meaning the last 10 or so difficulty increases) the reported network hash rate tends to peak/spike a day or two before the next difficulty increases, then drops off again - almost like some big players are trying to game the difficulty rate. The hash rate picks up again after the difficulty increase.

If that's really just statistics noise, there seems to be a fairly reliable pattern to it. But this time around the hash rate peaked around 2 days before the next difficulty adjustment and then kept dropping. It's back up to ~88 PH/s now.

I'd wondered about the same thing which is why I started to look at the stats. Certainly there have been spikes just before a difficulty change, but equally well there have also been ones earlier in the cycle and 4 changes ago the hash rate hardly changed for days at the end of the cycle.

I didn't publish the numbers but I did run simulations for the variations over 24 hours. At a constant hash rate we'll see a +10% spike every 10 days and a -10% spike every 10 too. Given that during our last difficulty change we had a 12.4% difficulty change then 10% at the end of the cycle is larger (in absolute PH/s) than 10% at the start.

I found the perception of a problem tends to be emphasised by looking at linear rather than log charts. The log charts show that the magnitude of the swings is pretty constant.

https://blockchain.info/charts/hash-rate?timespan=1year&showDataPoints=false&daysAverageString=1&show_header=true&scale=1&address=

With current mining hardware it seems unlikely that anyone would build a facility that would enable short term bursts of mining - as a strategy that wouldn't make a lot of sense. Doing this would incur a lot of expense in terms of power and infrastructure and if the aim was short term gain then the hardware would be run up to the point where the difficulty changes rather than dropping off a day or a few days before.

[davejh]

I'd wondered about the same thing which is why I started to look at the stats. Certainly there have been spikes just before a difficulty change, but equally well there have also been ones earlier in the cycle and 4 changes ago the hash rate hardly changed for days at the end of the cycle.

I didn't publish the numbers but I did run simulations for the variations over 24 hours. At a constant hash rate we'll see a +10% spike every 10 days and a -10% spike every 10 too. Given that during our last difficulty change we had a 12.4% difficulty change then 10% at the end of the cycle is larger (in absolute PH/s) than 10% at the start.

I found the perception of a problem tends to be emphasised by looking at linear rather than log charts. The log charts show that the magnitude of the swings is pretty constant.

https://blockchain.info/charts/hash-rate?timespan=1year&showDataPoints=false&daysAverageString=1&show_header=true&scale=1&address=

With current mining hardware it seems unlikely that anyone would build a facility that would enable short term bursts of mining - as a strategy that wouldn't make a lot of sense. Doing this would incur a lot of expense in terms of power and infrastructure and if the aim was short term gain then the hardware would be run up to the point where the difficulty changes rather than dropping off a day or a few days before.

Some follow-up analysis: http://hashingit.com/analysis/29-lies-damned-lies-and-bitcoin-difficulties

[ujka]

Some follow-up analysis: http://hashingit.com/analysis/29-lies-damned-lies-and-bitcoin-difficulties

A tip sent, for your article, time and effort to help us understand that, as you say: "what happens in the course of hour or even a few days; those numbers, tantalizing as they may seem, are largely meaningless. They are the lies among the truth that only becomes apparent over a much longer timescale.. "

[klondike_bar]

A  Big chunk of the profits being mined are re-invested into more equipment to offset the rising hashrate.  But the initial investment is not 100% recouped.

Mining bitcoins at home, for people who are capable only of low investments, is almost ended because of the small or null ROI they can have.

I disagree. at-home miners have a few big advantages:
1) free space (garage, basement, shed, etc). larger miners have rental locations and costs as such
2) "free" internet
3) 'free' wintertime heating (a large location will still be fighting to keep cool even in the winter)
4) no oversight or employee costs

IMO, mining at home can be more profitable than it is running a medium-sized farm. Ive personally had to move 8.5kW of my gear into a rental agreement to access enough power and space to run it outside the house. To go any bigger I am looking at a $500-1000/month rental elsewhere plus whatever costs are involved in adding AC/wiring/transformers to make it a useful space.

a lot also comes down to power costs, and anyone who pays less than $0.15/kWh has significant advantages over miners in many countries and US states

[davejh]

Some follow-up analysis: http://hashingit.com/analysis/29-lies-damned-lies-and-bitcoin-difficulties

A tip sent, for your article, time and effort to help us understand that, as you say: "what happens in the course of hour or even a few days; those numbers, tantalizing as they may seem, are largely meaningless. They are the lies among the truth that only becomes apparent over a much longer timescale.. "

Thanks 

[navigator]

A  Big chunk of the profits being mined are re-invested into more equipment to offset the rising hashrate.  But the initial investment is not 100% recouped.

Mining bitcoins at home, for people who are capable only of low investments, is almost ended because of the small or null ROI they can have.

I disagree. at-home miners have a few big advantages:
1) free space (garage, basement, shed, etc). larger miners have rental locations and costs as such
2) "free" internet
3) 'free' wintertime heating (a large location will still be fighting to keep cool even in the winter)
4) no oversight or employee costs

IMO, mining at home can be more profitable than it is running a medium-sized farm. Ive personally had to move 8.5kW of my gear into a rental agreement to access enough power and space to run it outside the house. To go any bigger I am looking at a $500-1000/month rental elsewhere plus whatever costs are involved in adding AC/wiring/transformers to make it a useful space.

a lot also comes down to power costs, and anyone who pays less than $0.15/kWh has significant advantages over miners in many countries and US states

At least 1 person understands...

[taipo]

The next leap looks to be a big-ish one...

Bigger than the last, most likely yes. But not biggish enough to be inbolded on https://bitcoinwisdom.com/bitcoin/difficulty

Just checked it today, seems to have dropped significantly in the past few days.

[64dimensions]

A  Big chunk of the profits being mined are re-invested into more equipment to offset the rising hashrate.  But the initial investment is not 100% recouped.

Mining bitcoins at home, for people who are capable only of low investments, is almost ended because of the small or null ROI they can have.

I disagree. at-home miners have a few big advantages:
1) free space (garage, basement, shed, etc). larger miners have rental locations and costs as such
2) "free" internet
3) 'free' wintertime heating (a large location will still be fighting to keep cool even in the winter)
4) no oversight or employee costs

IMO, mining at home can be more profitable than it is running a medium-sized farm. Ive personally had to move 8.5kW of my gear into a rental agreement to access enough power and space to run it outside the house. To go any bigger I am looking at a $500-1000/month rental elsewhere plus whatever costs are involved in adding AC/wiring/transformers to make it a useful space.

a lot also comes down to power costs, and anyone who pays less than $0.15/kWh has significant advantages over miners in many countries and US states

Totally agree with this.

1) In most places to rent and operate in a zoned business space you need a business licence.
2) Most times you have to register with the US state tax authority. I would have no idea if BTC yields count as revenue for tax purposes.
3) Doing any kind of electrical mods, of course you need permits. If you leave the space, you will have to pay to put the space back in it's original condition.
4) Most places since you are not a retail operation, this might be something that has to be in the correct industrial zone.
5) Space lease agreements want you to sign up for at least 5 years and some sort of liability insurance maybe required.
6) Finally, industrial settings could have pretty noisy power due to electric arc welders, rf generators and big inductive loads being switched on and off.

[Glizlack]

Until the new miners actually are released I still believe our jumps will be a bit lower than usual. I think for the next 2 months no huge increase. But thats just my opinion. When spoondolies and knc get their new miners out i think it will have a big jump but until then nothing unusual.

Steve