What do the larger farms know that I can't figure out?

[jtreble]

What do the larger farms know that I can't figure out?  With cheap power (USD 2-3 cents/kWh) I can't make a business case for the newest generation of miners.  All of my financial modelling shows the cheaper Antminer S9-series is still the way to go at the moment.  I know for a fact, however, that many of the larger farms in Canada are dumping the Antminer S9-series en masse.

What am I missing?

[Steamtyme]

The only thing you are missing is that the newer models are boasting efficiency. With cheap power that's not what you need at the moment. You should be piling in gear like s9's or m10's some of the Avalons fit nicely as well.. Really dense gear with mid level efficiency is you friend with low power.

Any large farm buying the nexgen of gear are doing do slowly. Phasing it in over months, as it makes sense to increase the hashrate while not breaking the bank for efficient gear that doesn't offer them the same bang for the buck against their electricity costs.

[philipma1957]

If you have paid off  10000 s9s making money at two cent power.

You add s17 and drop s9 do so by rolling back some profit into the new gear.

[wndsnb]

Farms could be taking the opportunity to sell their S9s before their price falls even further.

If you have 2.5 cent power, an S9 makes about $1.50 per day right now, but difficulty will be going up with all this new gear getting powered on. If we hit 120EH by the end of the year, that would cause diff to be 16.8T, and the S9 now makes about $1.00 a day.  Let's be generous and say we'll only hit 140EH by the halving. That means at the halving, diff will be around 19.6T, and the S9 is now breaking even. That is all assuming the price is still at $8300.

It would look pretty good if you can get $250 per S9 right now and start replacing them with gear that you are nearly certain will be profitable after the halving. In a month, you may only be able to get $100 for that S9.

[philipma1957]

Also if you are built out and paid off in s9 say 10000 sell 1000 and put in 100 s17+

If you get 270 for the s9 with a psu every ten you sell gets you a s17+

No money out of pocket. So you drop 130000th using 1200kwatts

You add 73000th using 320kwatts no money out of pocket.

This means you save almost 900 x .03 = 27 an hour or 27 x 720 = about 20000 a month in power cost.

If you have 100 s9s sell ten for 2800 and replace with 1 s17+

You drop 10x13= 130 th. Using 12000 watts.
Replace with 1 s17+.  You have 73th. Using 3200 watts.

Losing 47 th saving 9000 watts.

47 th earns ? 7.40 a day

9kwatts is 216kwatts a day at 5 cents that is $10.80 in power.
“.          “.      “                      4 cents that is $8.64 in power.

So it works for a medium farm.

With 100 s9s and 4 cent power or higher. Swapping out s9s for s17+ at about 10 pieces of s9 for 1 piece of s17+ makes sense.

[jtreble]

You should be piling in gear like ... m10's...

Thanks for this tip - the M10s looks interesting.  Strange that this model isn't listed on the MicroBT website any more.

[...]

Thanks for this - your "swapping" approach would also work for operators who have reached their power ceiling.  Interesting.

[Steamtyme]

Yeah finding an M10 is like a unicorn at the moment. They were a fairly short run of gear and they sold right out of them before launching the next gen as far as I could tell. When I bought my M20S I was actually hoping to pick up 2 or 3 M10's instead.

I think the power swapping thing is a very limited opportunity. Anyone really wanting to take part in that should more or less undercut and hope for 200 USD. There are so many pallets of used S9's just sitting out there waiting. Then the more you sell the less people need them. Not always the best but people can at least get some cash back on the gear, which is like a coupon for the upgrade. Every day this deal becomes less and less likely, as you are really trying to sell to people in the same boat as yourself power cost-wise.

[mikeywith]

I don't really own a "large" farm , but I do have a good amount of mining gears and I have free power (not even 2 to 3 cents as per your question), but I have a few reasons on why I want to get rid of my S9s pretty soon and replace them with more efficient gears.

1- The set-up cost (building/rent,wiring,racks,switches,cooling)

While I don't pay for power, I still do have other expenses that I need to pay for, so once S9s become barely profitable to be worth maintaining in the first place , I will have two options to save my business

A- Get more S9s :

This will not only cost me more money in gears, but it will cost me more in the infrastructure from wiring and what not, and i rather save the money I have to spend on that, to get more efficient gears now.

B-Sell my S9s and buy more efficient gears ( no infrastructure expenses required)

But honestly, if someone with free power gets to the point where having say 50 S9s is not worth keeping, how much will anybody else pay me? 10$ each?

2- The miner's lifespan

Many S9s that I own are bough used, I have no clue how many months/years the Chinese mined with them, and they are likely to fail sooner than any new gear,replacing them now with brand new gears will keep me in the game a bit longer.

3-Limited power ( does not apply to everyone)

To me, the free power I have is limited, and I rather maximize it with more efficient gears.

Having said that, it does not mean I am going to jump in right to S17 or S17 pro, because then I won't be fully utilizing the free power I have , but gears like T17 or Whatsminer M21s seem pretty reasonable for a balance between efficiency and price per TH.

[Huxley23]

I have a similar situation as you. But honestly, I do not see a point in buying a more efficient gear with mined BTC. There is no way, the new gear will ever earn me the same amount of BTC that I had to pay for it. Therefore I am waiting for post halving prices and I might buy some new S17.

[Philipma1957cellphone]

We will be adding. 200th from oct 24 to nov 14

[mikeywith]

I do not see a point in buying a more efficient gear with mined BTC.

Spending money = spending money, it does not matter if you buy the gear using your mined btc or you sell your kidney for it, also the fact that you will never get your btc back is not certain, if btc price was to drop, then it would be a better option to have bought a miner than simply hodl'ed bitcoin, the best would be keeping fiat in that case, but if you do that and prices spike, you will have to pay more for gears and well as btc, so really it's a game of luck, so  balancing between keeping cash, btc and slowly upgrading to more efficient gears is the way I like to play this game.