Difficulty and Halving Speculation.

[mikeywith]

We are about 350 days to BTC halving, hanging at all time high difficulty 7.46T.

I understand that in normal conditions, we are suppose to have double , tipple or even 100 times the difficulty a year from now, but with the halving around the corner, I seem to think that the situation might be a bit different.

Bitcoin price as well as the the improvement in mining technology will indeed have a huge impact , and for the sake of simple speculation, let is just pretend that before the halving the most efficient gear out there will do 100th at 2000w ( not very well though of but seems pretty realistic) , and given the slow production of this new gears while assuming that BTC will hit a new ATH.

I would like to know what do you guys think about the effect of the rewards halving in terms of difficulty ? what do you "guess" the difficulty will be by then?

[philipma1957]

I think that a 100th machine at 2000 watts is not coming before the 1/2 ing.

The one year guess for diff is tough.

I guess price first about 15k to 20k.

I guess network efficiency will be 50 watts a th.

Now to go for the diff.  At 50 watts a th. And 15k a coin and 8 cent power.

Wait for the number.  Diff could be 21t and you turn profit.

The s17 is about 40 watts a th so for the network to change from 90 watts to 50 watts a th needs to happen and will happen. Since there is lots and lots of spare s9 gear getting the network to drop from 90 to 50:watts a th will be a slow process.

In fact diff is look to drop back to 7.2 th next jump.

I think people do not realize the s9 on the network  will need a long time to go away. Maybe a full year.

[mikeywith]

@philipma1957 thanks for the great analysis as always, but any particular reason why do you think a whole year is not enough to double up on the efficiency of current gear ( S17) ?

Code:

Model   - TH    - Power - Date                     Efficiency 

S11     - 20     - 1500w  - Sep 2018              75
S15     - 28     - 1500w  - Dec 2018              53  
S17     - 56     - 2500w  - Apr 2019              44  ( factor of 1.7)

S19     -------      - ------   - ---------              ----------
S21
S23     -100      -2500W   -Oct 2019           25 ( factor of 1.7)

The Oct 2019 is merely a guess based on the  7 months from S11 to S17 and 1.7 factor, but still , even if almost double the time , say 12 months or almost right after the halving we could very likely see an S23 with 100th and 2500w.  no? maybe? yes? noway?

**The above coded part is not from bitmain, just my own speculation.

[philipma1957]

In october 2018  we were at   45,000,000,000 gh most of that was  s-9s   most was set to 100 watt not 80 watt.

we are back to 47,000,000,000 gh  it is a blend  but I am sure a lot of it is still  s9  the s9 still makes money for a 4 cent miner.

it will take time and diff jumps to force it out,   if price drifts up to 10-12k  it will stay longer.

a of  diff  7.77 coins at 7793 power at 4 cent the s9 makes 2 dollars daily set to 13th and 1300 watts it is viable to mine it till it breaks.

If I have 100 of them  they make 200 a day.  If I can get the s17 for 2400  I  earn one every 12 days. I add it in.  and set 7 of my s9's to 10000 th and 820 watts using braiins. that means my watts drop 7 x 480 = 3360 the s17 is 2360 watts so I have a net loss of 100 watts.

hash drops 21 th and the s17 is 55 th I gain  34th  . 100/7 = 14 times  I can do this every 12 days  so in    14 x 12 = 168 days time.

all my 100 s-9s are on line  set to 10th and 820 watts.

have 14 s17's on line.  each time I did this I saved 1000 watts  so I now have 14 x 1000 = 14,000 watts or 6 more s17's.

same power footprint.

100 s9's set to 820 watts     100 x 10th = 1000th  at 82000 watts.

20 s17's using 2360 watts     20 x 53th = 1060th at  47200 watts.

so   2060th at  129200 watts = 62.7 watts.

so dropping the network to under 50 watts  a th  is hard unless every s9 and a841 disappear.

Why should they  stop  mining if they earn 2 bucks a day  as they do right now for a 4 cent power guy.

bitmain has no $$$ need to release a s23  so they won't.  The s17 is leaps and bounds better then the s9  and the s9 will need at least 18 months to leave mining based on 4 cent power.  since at a coin price of 7800 the s9 turns a profit at a diff of17,7th.

[mikeywith]

so i am right ( at least not wrong ) in terms of the technical limitation, but looking at it from an economical point of view as you did, it does not make sense for bitmain to go for s23 while the S9 is still selling easily, but come to think of it.

1- The same thing for s15 , it was out of stock, bitmain could easily sell whatever they make, but they still went for the S17
2- The market competition in terms of making the most efficient gear at any give time.

based on the 2 points above , i see no reason why bitmain wouldn't push towards s19 / s21 and eventually s23 in a year or so , unless i am missing something in the equation.

[philipma1957]

No you are not missing but last year they flooded out the market with s9's  Say April 1 to Sept 1.

They blew through the s11 and the s15 to get to the s17  which can do 38 freaking watts a th  truly amazing.

No one can do a lot better then 46 or 47 watts a th.  So they are back in the saddle  with a clear cut edge over every one.

They  won't push for  the  s19 s21 s23  until right before or after the ½ ing.  They simply do not need better gear when they are the best gear  by  a  15-20% efficiency margin.

Also  the m20 and its like may have an issue with 3417 watts on a c19-c20 cable.

the s17  with 2 cables is better design.

I do not see them doing any new gear until april 2020 at best.

[mikeywith]

I am convinced by this theory, so far the S17 pro is pretty a head of every other gear in the market in terms of efficiency, Avalon does not seem to be giving Bitmain a hard time so i think it's safe to exclude them from the equation , INNO is probably the only real competitor to Bitmain , so Bitmain pushing for s19/s21/s23 will pretty much depend on how fast inno moves forward with their next gear.

all of this is based on the fact that these guys don't work together in terms of delaying the release of new gears and what not, after all there is a very good chance that all of them are on the same line somehow.

[philipma1957]

I am convinced by this theory, so far the S17 pro is pretty a head of every other gear in the market in terms of efficiency, Avalon does not seem to be giving Bitmain a hard time so i think it's safe to exclude them from the equation , INNO is probably the only real competitor to Bitmain , so Bitmain pushing for s19/s21/s23 will pretty much depend on how fast inno moves forward with their next gear.

all of this is based on the fact that these guys don't work together in terms of delaying the release of new gears and what not, after all there is a very good chance that all of them are on the same line somehow.

Well bitmain’s s17 can do 38watts a th that is a big money maker.

Bitmain has 4 cent power and they would be crazy to not have built some solar fields as well.

4 cent power is about 2.50 usd a day for this machine which is earning  0.00003520 for 1 th.

That is  0.00176 btc a day or 12.50 usd so it makes 10 dollars a day for them.

They preordered out until sept.  Which means. They sold the machine twice.

Ie they build it with your preorder money then mine it till it earns what you already paid them.

It is why I really resist long preorders as much as possible.

You get a 95 day old machine they dip,clean and swap the fans.

So those 1700 usd preorders make more like 2700 for them.