2023 Diff thread now opened.

[DaveF]

Was talking with a friend over dinner the other day and he thinks there are going to be a few big difficulty jumps up early in the year followed by a long period of flat.

His logic more or less is that with all the bad news coming out of the industry money people will be less willing to invest so large new mining farms are not going to come online but those that are surviving will keep going. Some will drop off, their gear will be sold and keep mining. Even places will secured funding are going to be more hesitant to buy since intersect rates are up and that is going to make them not want to spend.

Small individuals will keep mining some will get more efficient gear others will leave but in the end that is not going to matter much.

But what he thinks is going to be the biggest diff movers is power. A lot of places that got cheap rates in 2020 / 2021 for 2 and 3 year contracts are going to be coming to an end this year. will they be able to keep them or are they going to get a penny or two per kwh that takes them from small profit to small loss. And they just pack it in.

The company he works for is looking at that issue now. They had an OK but not great rate for power at their manufacturing plant and it's going up just a bit. But as the owner of the company said to him, that bit is going to be close to $3k a week during the summer when the chillers are running all day. Less in the winter but it's still probably going to cost them close to an extra $70k to $100k in power this year. Which considering their overall cost is not much. But it means that they are not getting that new machine they wanted.

Miners are going to be in the same hurt unless they planned for it.

-Dave

[1715669138]

1715669138

[1715669138]

1715669138

[1715669138]

1715669138

[1715669138]

1715669138

[1715669138]

1715669138

[philipma1957]

Was talking with a friend over dinner the other day and he thinks there are going to be a few big difficulty jumps up early in the year followed by a long period of flat.

His logic more or less is that with all the bad news coming out of the industry money people will be less willing to invest so large new mining farms are not going to come online but those that are surviving will keep going. Some will drop off, their gear will be sold and keep mining. Even places will secured funding are going to be more hesitant to buy since intersect rates are up and that is going to make them not want to spend.

Small individuals will keep mining some will get more efficient gear others will leave but in the end that is not going to matter much.

But what he thinks is going to be the biggest diff movers is power. A lot of places that got cheap rates in 2020 / 2021 for 2 and 3 year contracts are going to be coming to an end this year. will they be able to keep them or are they going to get a penny or two per kwh that takes them from small profit to small loss. And they just pack it in.

The company he works for is looking at that issue now. They had an OK but not great rate for power at their manufacturing plant and it's going up just a bit. But as the owner of the company said to him, that bit is going to be close to $3k a week during the summer when the chillers are running all day. Less in the winter but it's still probably going to cost them close to an extra $70k to $100k in power this year. Which considering their overall cost is not much. But it means that they are not getting that new machine they wanted.

Miners are going to be in the same hurt unless they planned for it.

-Dave

Well a strong diff move upwards and a few days later a 7% price rise.

[mikeywith]

Yep lets argue you have 100000 s19 pro they do 3kwatts and 100th or 2kwatts and 82th

flip flop in a few minutes when needed will be the norm.

So you are burning 300000 kwatts a hour and 10eh or 200000 kwatts an hour and 8.2 eh

now the move above would be a 300mw mining farm and I am not sure if one does exist at that size.

Keep in mind that 100000 S19 pros don't have to be in the same farm, they can be in a thousand different farms, they don't have to be in the same location, they just need to have a similar power rate to act similarly.

@Dave, everything is possible, but a few large jumps at current prices are unlikely to be sustained, we are talking about the latest generation gears at 5 cents, let alone the other fewer effienct gears, with the coming difficulty jump life will be a lot more difficult for those who have the latest gear + below average power rate, even if someone is willing to invest more money, someone else will have to leave, without a good price increase I don't see how a few large diff jumps can happen or even if they happen, not sure how can they be sustained.

But this market is crazy, you can only tell for sure when time has passed.

[philipma1957]

Yep lets argue you have 100000 s19 pro they do 3kwatts and 100th or 2kwatts and 82th

flip flop in a few minutes when needed will be the norm.

So you are burning 300000 kwatts a hour and 10eh or 200000 kwatts an hour and 8.2 eh

now the move above would be a 300mw mining farm and I am not sure if one does exist at that size.

Keep in mind that 100000 S19 pros don't have to be in the same farm, they can be in a thousand different farms, they don't have to be in the same location, they just need to have a similar power rate to act similarly.

@Dave, everything is possible, but a few large jumps at current prices are unlikely to be sustained, we are talking about the latest generation gears at 5 cents, let alone the other fewer effienct gears, with the coming difficulty jump life will be a lot more difficult for those who have the latest gear + below average power rate, even if someone is willing to invest more money, someone else will have to leave, without a good price increase I don't see how a few large diff jumps can happen or even if they happen, not sure how can they be sustained.

But this market is crazy, you can only tell for sure when time has passed.

we are near the sustainability level of price and diff for big players.

we may get bump up bump down for a while.

in this zone diff leads price which most people think wtf.

but strong downward move in diff and prices dragged.

this jump strong move up in hash and prices lifted.

It is subtle but a bit different then the other way.

Now we could get strong dropping diff five or six jumps in a row with a price drag just behind it.

Or maybe a nice set of plus diff and price going upwards.

I see it more as a see saw coming.

still hopefull.

[mikeywith]

in this zone diff leads price which most people think wtf.

but strong downward move in diff and prices dragged.

this jump strong move up in hash and prices lifted.

I don't really think diff lead prices directly, some investors might use the hashrate as an indicator for the overall strength of the BTC ecosystem and inject more money when they see miners are building and are confident, but no direct impact whatsoever in my opinion.

In fact, even the other more popular claim that price leads diff isn't all that accurate, people can say whatever they want for almost a year now we have been going down in price and up and in diff, this happened before as well, if 1 year isn't long enough to arrive to the conclusion that price doesn't "directly" affect hashrate then I don't know what is!

I think the most accurate word to be used is "profitability" which has a lot more stuff in it than just the price direction, any gains in gear efficiency will cause a diff spike, any drop in power prices will cause it, and more lenders willing to give money to miners will lead to diff jumps, any drop in gear manufacturing cost will lead to diff spike, and of course, the opposite is also correct.

The only time when diff-price start to have a direct correlation while ignoring everything else is when a defined equilibrium is reached by all miners, you have x number of miners, x of them pay x $ per KhW, and none of them can afford to add new gear, none of them is forced to sell/shutdown, some of them are mining at almost breakeven, at this stage (which is very difficult to be at) you will see a very direct and fast impact caused by price change, a 2% drop in price will most likely to a 2% drop in hashrate, and vice versa.

Without being in such a status, we don't know what we don't know, we don't know if someone figured out a way to access 5000MW at 1 cent and has 10 billion $ in the case and is going to go all in, such an operation can still happen even if the price was to take another dive to 10k, so most people assume a drop to 10k means 100% a huge drop in difficulty which isn't true, it might be true for your own operation that 10k isn't profitable for you, it could still be profitable for 90% of the other miners.

With that said, and given that it's extremely unlikely for anyone to find a 5000KW at 1 cent, we know a 50% increase in hashrate while price is 17k is very, very unlikely, even if said person does manage to find it, we know many others pay 5-6 cents and will be forced to shut down if this 5GW guy fires his gears,  we just don't know what a 50% increase in hashrate will exactly mean to every individual miner out there, will 25% of them be mining at loss? maybe only 10% of them will, so maybe 100EH in vs 10EH out, the other miners will still mine on just with a lot less profit.

So what we try to do here isn't something you can take to the bank or include in a scientific research paper, but we know the average power rate for the large guys at least is anywhere between 3-5 cents, and their gear efficiency is between 25-35w/th, we know that at 10k BTC most of them will be mining at a loss, we know that a 50% drop aka 50% drop in profitability isn't something they can take so it either doesn't happen at all or does happen only to be wiped out but whoever decides to leave first.

But all these assumptions we make are subject to being very inaccurate.

[stompix]

But this market is crazy, you can only tell for sure when time has passed.

Best conclusion!
We barely had a small change in price that probably shouldn't be enough to put more than to s5 back online and here we have a 109.9106% pace, and with a  current average of 6 cents per TH here comes another maybe 10% dif on, to make things really spicy.
Every single thing, from electricity prices to bitcoin prices, news of miner troubles, mining companies' stocks getting hammered and asking for a loan over loan or even planning on shutting down points to a final peak in difficulty, but no, let's have another ATH because things are so great, and let's have another ATH because some have already bought miners with delivery in Q4 2023!

And it all makes as much sense as Riot shares going up 50% from ~3.30 to ~4.90 mainly because they've announced a name change from Riot Blockchain to Riot Platforms! Of course, not so much joy for the ones that bought at 70$ two years ago, but that's a different story!

In fact, even the other more popular claim that price leads diff isn't all that accurate, people can say whatever they want for almost a year now we have been going down in price and up and in diff, this happened before as well, if 1 year isn't long enough to arrive to the conclusion that price doesn't "directly" affect hashrate then I don't know what is!

I still think that price leads the difficulty but starts to become less accurate when the profitability you mentioned has been at through-the-roof levels, it takes a ton of gear to level those to keep the same balance as before and all this gear is almost impossible to be even produced in a few months not even mentioning actually being shipped and installed.
If Bitcoin would have been only CPU minable probably the spikes and the correlation to the price would have been far more accurate, people could have just bought ready available processors on the market in the millions or used the immense amount of computing power available to rent already. With ASICs, by the time some that have been ordered when the price was going up are getting installed when the price is 1/3 because not doing so means you lose more money! The same person who is now turning on that gear would have never bought it right now, but it's a done deal for them.

So with readily available miners, we could have seen 400Exash back in late 2021 and then a downslide to a bit above current levels now.
But again, just an assumption!

[philipma1957]

But this market is crazy, you can only tell for sure when time has passed.

Best conclusion!
We barely had a small change in price that probably shouldn't be enough to put more than to s5 back online and here we have a 109.9106% pace, and with a  current average of 6 cents per TH here comes another maybe 10% dif on, to make things really spicy.
Every single thing, from electricity prices to bitcoin prices, news of miner troubles, mining companies' stocks getting hammered and asking for a loan over loan or even planning on shutting down points to a final peak in difficulty, but no, let's have another ATH because things are so great, and let's have another ATH because some have already bought miners with delivery in Q4 2023!

And it all makes as much sense as Riot shares going up 50% from ~3.30 to ~4.90 mainly because they've announced a name change from Riot Blockchain to Riot Platforms! Of course, not so much joy for the ones that bought at 70$ two years ago, but that's a different story!

In fact, even the other more popular claim that price leads diff isn't all that accurate, people can say whatever they want for almost a year now we have been going down in price and up and in diff, this happened before as well, if 1 year isn't long enough to arrive to the conclusion that price doesn't "directly" affect hashrate then I don't know what is!

I still think that price leads the difficulty but starts to become less accurate when the profitability you mentioned has been at through-the-roof levels, it takes a ton of gear to level those to keep the same balance as before and all this gear is almost impossible to be even produced in a few months not even mentioning actually being shipped and installed.
If Bitcoin would have been only CPU minable probably the spikes and the correlation to the price would have been far more accurate, people could have just bought ready available processors on the market in the millions or used the immense amount of computing power available to rent already. With ASICs, by the time some that have been ordered when the price was going up are getting installed when the price is 1/3 because not doing so means you lose more money! The same person who is now turning on that gear would have never bought it right now, but it's a done deal for them.

So with readily available miners, we could have seen 400Exash back in late 2021 and then a downslide to a bit above current levels now.
But again, just an assumption!

price is usually the leader not difficulty.

mostly because price can go so high so fast it takes difficulty a year or so to catch up.

sept 2017 diff was 1.1t
dec 2017  diff was 1.8t


yet price went from 4600 to 19800

the diff needed to gain way more and it moved all the way up to 7.4t on oct 2018.

it over corrected because the gear takes too long to show up and mines kept ordering past the point of correction.

we got diff correction  oct 2018 to may 2019

7 t went down and took till may to get back to 7t.


in the next cycle China fucked it all up by just cutting the diff in have.

but

July 2021 we were 29000
 Nov 2021 we were 68000

diff did
July 2021 13.9t
Nov 2021 23.0t


we got to

30t may 2022


and instead of China fucking things up we had Solana crashing

and diff dropped .

But I feel we are very close to the October 2018 time span.

Of course outside shit like

SBF/FTX
Solana
China cutting hashrate

all interfere with the diff. and price.

it is simply not diff+price+power= profit.


new miner below really undercuts the s19 xp

122th for 2379

https://shop.bitmain.com/product/detail?pid=00020230108213609854b369SGwI0654

[NotFuzzyWarm]

122th for 2379
https://shop.bitmain.com/product/detail?pid=00020230108213609854b369SGwI0654

As expected, to keep the 3-3.5kw 'sweet spot' I bet all they did is pile in more chips and use their lower power FW to do it...

[mikeywith]

122th for 2379
https://shop.bitmain.com/product/detail?pid=00020230108213609854b369SGwI0654

As expected, to keep the 3-3.5kw 'sweet spot' I bet all they did is pile in more chips and use their lower power FW to do it...

And no more 2 wires supplying the PSU, so now a single 20A socket is a must for this model, they seem to be catching up with the custom PSUs that some folks like Alpha make, supporting the 277V, many U.S based factories or industrial business, in general, have 480V three-phase wye wiring coming in the building, so they lowest they can go with Phase to Neutral is 277V, the previous models were not compatible with such voltage and people needed to replace the original PSU.

These are just more signs that Bitmain and the others don't give a flying fudge about home miners anymore, maybe it's about time for those waiting for a 500-1000w home miner from Bitmain to give up.

[philipma1957]

122th for 2379
https://shop.bitmain.com/product/detail?pid=00020230108213609854b369SGwI0654

As expected, to keep the 3-3.5kw 'sweet spot' I bet all they did is pile in more chips and use their lower power FW to do it...

And no more 2 wires supplying the PSU, so now a single 20A socket is a must for this model, they seem to be catching up with the custom PSUs that some folks like Alpha make, supporting the 277V, many U.S based factories or industrial business, in general, have 480V three-phase wye wiring coming in the building, so they lowest they can go with Phase to Neutral is 277V, the previous models were not compatible with such voltage and people needed to replace the original PSU.

These are just more signs that Bitmain and the others don't give a flying fudge about home miners anymore, maybe it's about time for those waiting for a 500-1000w home miner from Bitmain to give up.

our 320kva transformer will work well with this.

we will be adding some of these soon.

[NotFuzzyWarm]

I just took a quick look at their ad and  missed the higher input voltage. Ja, good move for them and most large farms but of course it also screws folks running 200-208v  for their main power

btw: its max current is spec'd at 25A, not 20. What I do *not* see is a spec for what that input power connector is. BM says that is a locking connector but thass all. Heavens forbid they tell us so folks can get the right one when making the no-doubt custom cords needed.

[mikeywith]

but of course it also screws folks running 200-208v  for their main power

Not really, the operation input voltage is stated in their specs as well (200~277v), I don't know what do they mean by 25A, probably the power cord is able to carry 25A, or they are just being extra careful to not take any legal responsibility, but 20A is more than enough even accounting for the lowest voltage of 200V + 20% buffer gets you to 20.13A, if I were to buy one of these I would comfortably run them on 20A, I run some power-hungry Whatsminer that suck nearly 3500w, sometimes when the demand on power is so high the voltage drops to 210v and stays there for a couple of hours,  it puts me at 16.6A, on a 20A/socket PDU, I have not had any issues.

Also, most PDUs out there (the ones for mining at least) will come 20A sockets, even the largest mining PDU I saw is 160A per phase, with 8*20A ports on each phase, this is probably one of the reasons why Bitmain, Whatsminer, Avalon and etc, are not going past 3500W per miner, actually going past that on a single phase is just asking for troubles, it would require the re-making/re-building of so many stuff.

[kano]

Meanwhile, if the current trend continues for the last 4 days of this Diff (on topic )
that's over 20EH come online in the past 2 weeks.
Interesting where 200000 100TH/s miners came from and where they got their 500MW from
(assuming a low 25J/TH)

[NotFuzzyWarm]

Not really, the operation input voltage is stated in their specs as well (200~277v), I don't know what do they mean by 25A, probably the power cord is able to carry 25A, or they are just being extra careful to not take any legal responsibility

Ja, most odd that their large ad at the top says 220-277VAC and yet the detail specs say 200-277V. My guess is that for running on a lower line you have to run at reduced speed.

[mikeywith]

Ja, most odd that their large ad at the top says 220-277VAC and yet the detail specs say 200-277V.

Bitmain.

My guess is that for running on a lower line you have to run at reduced speed.

I can't tell for sure, but I know the other models which I either owned personally before divorcing Bitmain or installed to some clients did run well at full speed at as low as 200V, they probably do factor that in their PSU specs, in fact, I have not seen any electronic piece that wouldn't run just fine at 200v, in many countries (probably 50 or so), the 3P is actually just 380v and not 400V.

Where I live, for most places, the single-phase voltage would be in the 210v range at best, those living pretty close to the transformer/station might get 220v or slightly higher, the further you go the lower it gets, all the way down to 200v, sometimes even as low as 195v when demand on power is high the for some reason the power company doesn't touch the tap changer.

I know 200v for those who have 240v might seem pretty low, but it's pretty much normal, there is nothing fancy about any PSU running at full load on 200v, I know the S19 pro works fine at full speed at 200-210v, so does the S19, S17 and a few other models, so my guess this one will not be different.

that's over 20EH come online in the past 2 weeks.

Price went up 8% this epoch so far, natural gas prices are down like 50% this month so electricity prices must have dropped a good amount for many miners, so the majority of those 20EH could be old gears that were either offline for a while or were running at lower speeds, with the addition to some previous large orders from Bitmain are being delivered throughout the year.

[stompix]

These are just more signs that Bitmain and the others don't give a flying fudge about home miners anymore, maybe it's about time for those waiting for a 500-1000w home miner from Bitmain to give up.

The writings were on the wall for a long time, it is clear they don't want to design miners for hobbyists, small miners and it seems like they don't even want to deal with individuals anymore, tolerating them just because they bring a few more sales when the market is down but I expect this to change and go to MOQ orders far more often than now.

I'm curious, if somebody has a clue, hunch, idea, how much more would 6x500W miners costs compared to 1x3000W, assuming the same numbers of chips are used overall? Probably the case doesn't matter at all here since I managed to make a serious bump on one by dropping a small plier from like 90 cm on top of one.

Price went up 8% this epoch so far, natural gas prices are down like 50% this month so electricity prices must have dropped a good amount for many miners, so the majority of those 20EH could be old gears that were either offline for a while or were running at lower speeds, with the addition to some previous large orders from Bitmain are being delivered throughout the year.

US inflation data is incoming, the last month it was already counted in as the price moved before the release, same as now, in the previous two months the price jumped exactly when the numbers were released, if we have a combination of the two and another spike probably every miner who kept his gear around will turn them on, also a lot of small miners who will think of mining as acquiring bitcoin cheaper before a new FOMO.

But I don't see another 10% jump next period, so, based on my prediction record 

[philipma1957]

CPI is in

https://www.usatoday.com/story/money/2023/01/12/cpi-report-data-release-gas-prices-inflation-december-2022/11034854002/

CPI report shows fall in gas prices helps inflation slow again in December ...

Inflation eased substantially for a third month in December as tumbling gasoline prices and a moderating rise in grocery bills offset another surge in rent.

Consumer prices increased 6.5% from a year earlier, down from 7.1% in November and a 40-year high of 9.1% in June, according to the Labor Department’s consumer price index..

That marks the slowest annual gain since October 2021 and matches economists' estimates.

On a monthly basis, consumer prices fell 0.1%, the first monthdecline since May 2020.

What does core CPI mean?

Core prices, which exclude volatile food and energy items and better reflect longer-term trends, increased 0.3% from November following a 0.2% bump the previous month. That lowered the annual increase from 6% to 5.7%.

that -.1% is a pivot

very likely we see .25 rate jump and not .50



so price rallied and diff went under +10%

https://www.bitrawr.com/difficulty-estimator


Latest Block:   771645  (24 minutes ago)

Current Pace:   108.7197%  (1534 / 1410.97 expected, 123.03 ahead)

Previous Difficulty:   35364065900457.12                           
Current Difficulty:   34093570325203.84                           
Next Difficulty:   between 36965242588477 and 37154981200221
Next Difficulty Change:   between +8.4229% and +8.9794%
Previous Retarget:   January 2, 2023 at 10:49 PM  (-3.5926%)
Next Retarget (earliest):   Sunday at 7:52 PM  (in 3d 1h 53m 25s)
Next Retarget (latest):   Sunday at 9:24 PM  (in 3d 3h 25m 50s)
Projected Epoch Length:   between 12d 21h 3m 6s and 12d 22h 35m 31s
Copy stats to clipboard

so if we end at +8% diff

and 18,100 price both will be up 8%

or sideways

[kano]

Heh well it was 9% the other day and still over 9% today - so as usual their range is out
Currently it's +9.33%
But who knows, another almost 3 days to go.

[altair_tech]

Price bump brings in more hash.
Probably going to end up with a double digit difficulty rise.
This is why we can't have nice things

[philipma1957]

Price bump brings in more hash.
Probably going to end up with a double digit difficulty rise.
This is why we can't have nice things

Actually the diff is a bit less than it was earlier it was over 11%

this indicates that

A) most gear is now online
B) next jump will be far lower than this one.

this jumps comes in 2-3 days