I'm going to take a stab at repairing an S17pro I have with a dead dashboard, so I'm looking for another dead dashboard, or a dead S17pro to use for parts.

Well if the price is good enough I'm sure there would be a lot of people who would like to convert their S9 paperweights into something useful. Break-even for 55W/Th is about 5.6c/kWh. Could work if the price is around $60 per board. Of course if BTC tanks you'd be left holding the bag unless you got commitments before ordering.

OP said a rack only fits 4 S9s, so the $30 month fee per rack kills current profit on a S9 or L3+ I think.

4 S9s on enhanced low power mode doing 80W/th would be around 40TH for 3200W.

40TH 21 hours a day would produce .00868BTC per month, currently about $79.

3200W, 21hours a day would cost  $51 for power + $30 maintenance, so $81 per month.

OP, can you post a photo of a rack?

Would probably work well for Avalon 1066s. Would 3 Avalon 1047s fit in a rack? A 1047 is  190mm x 190mm x 292mm.

---

2 1066s would earn 0.0216BTC per month and would cost $134, so about $63 per month profit.
3 1047s would earn 0.024BTC per month and would cost $144, so about $75 per month profit.

Looking to sell two Avalon 1047, 37TH @ 2380W. Been running since October in a hosting facility in upstate NY.

Trusted forum escrow is welcome.

Asking $400 each plus shipping plus escrow fee. Payment in BTC.

PM me your location so I can check the shipping cost. Maybe we can make a deal.

Looking to sell two Avalon 1066, 50TH @ 3250W. Been running since January in a hosting facility in Labrador, Canada.

Trusted forum escrow is welcome.

Asking $550 each plus shipping plus escrow fee. Payment in BTC.

I suspect there may be differences between different firmware versions. I attempted to use the API to set the pool on a few 1066s and ended up with two miners that are hashing nowhere. Been waiting 2 days for my host to do a factory reset to get them back up... The web UI freezes up after the login screen.

Some of the commands listed in the API returned an unknown command error. But the reboot, idle, and pool commands didn't return any errors.

Has anyone tested out the power consumption of the latest 1066 and 1047 firmware? The release notes say they optimized power consumption.

https://canaan.io/service/447

Also, has anyone played with the "Privileged Firmware"?

https://github.com/Canaan-Creative/avalon10-docs/blob/master/Privileged%20API/Previleged%20API%20USER%20MANUAL%20v0.6.pdf

update: I had my host test out my 1066s with the updated firmware and the power consumption came out to 65W/th in both "High Performance" and "Normal" modes. No efficiency gain for normal mode.

Yup, PV=nRT, ideal gas law. If you increase the pressure of air and hold the volume constant, then the temperature will go up. When you lower the pressure, the temperature goes back down.

This can help in the situation you are describing because when you increase the temperature of your incoming air it is causing a larger temperature difference between the air and the ground, increasing the amount of heat you can transfer. When the pressure lowers on the other side, you end up with cooler air. But the heat still needs to go somewhere, just increasing and then decreasing the pressure does not make the heat disappear. You also need to put energy in to increase the pressure in the form of some fans that can push a very high static pressure.

I'd guess that Phil is correct about a flow-through system vs a closed loop, but it does depend on exactly what the temperature rise is through the miner. There is some max input temperature that the miner can operate at, and at whatever that temperature is, the exhaust temperature would need to be higher than ambient for a flow-through system to beat out a closed loop. This would probably only be an issue on the hottest days of the year.

Shading the input is going to make 0 difference especially when you're pulling several thousand CFM though. It may feel cooler in the shade, but the air temperature isn't actually lower. I guess it could make a measurable difference on a windless day if your intake was in the middle of an asphalt parking lot...

Dry soil has a pretty low thermal conductivity which will limit the amount of heat you can extract from your incoming air. So you may need hundreds of feet of pipe per miner to operate on the hottest days of the year.

So let's say the maximum input temperature a miner will work at is 90 deg F at full fan speed of 200cfm. On a 105deg day, it would require 3240btu/hr to take 200cfm of air down to 90 deg, so if the efficiency is similar to a liquid system you'd need 324 feet of pipe per miner. If the temp on the hottest day of the year was 115deg, then you'd need 5400btu/hr or around 540ft of pipe per miner. That might be the case for a 1200W S9.  Newer generation miners use more air, so numbers would be larger.

btu/hr = 1.08 * Temperature Drop * CFM

Agreed, if we get a 50% inflation rate for the period of a single hosting contract, and the host doesn't have a flat contract rate with his energy provider, then that would be an issue for the host.

Yeah, I saw articles on systems like you are describing. It's trading off efficiency for simplicity. It's just a lot more efficient to transfer the heat from a liquid to the surroundings than from air, and liquid holds way more heat per unit volume so you don't have to move it as fast.  Keep in mind that you'll need to push around 200cfm of air per miner through your pipe, so you'll need very large pipes and an air system that can handle very big static pressures. So for the scale you're talking about, I think the liquid one with a heat exchanger is probably going to end up simpler in the end. But in my opinion, neither will be as inexpensive or as simple as an evaporative cooling system.

But I'd love to see the results of a test. Just set up a single air loop with a single miner, and bury 50 or 100 feet of 6 or 8" pvc.

That I agree with. Inflation is irrelevant and was more of the point I was trying to make.

An interesting fact is that nearly 40% of electricity in the USA is generated from natural gas. If the price of natural gas spikes, the price of electricity on the grid will follow. But the market price of gas may not have much impact the OPs situation if he is using waste gas or is buying directly from the producer, bypassing distribution costs.

But we're getting off-topic here... OPs question was about what price would attract miners, not arguing about aspects of his situation that he is likely the expert on.

For that question, I'd say that this probably isn't the best time to ask with the halving only a few weeks ago and the diff not even stabilized yet. But the market price in North America at the moment seems to be in the 6.5c range for small quantities, closer to 5c for larger quantities. The trend in prices have been down, a year ago the prices available were higher and hosting operations that can't afford to go any lower are going out of business. It wasn't long ago when it was hard to find hosting for small quantities below 10c/kwh. Race to the bottom.

Like everyone else already said, not easy.

I'd have to disagree with most of what HardFacts said though. You only pay taxes on profit, running a mining operation is not labor-intensive, and if gas cost is going to go up because of inflation, then the wouldn't the price of gas in BTC not be affected....

The market price of natural gas can be volatile, but it seems like the OP might have access directly to a well, so his price may not be tied to the market price.

Scaling up can be very difficult. Big difference between setting up 20 miners and 500 miners. You'll need an electrician that has experience with datacenters. A building (or shipping container) full of miners generate harmonic currents that can cause very large neutral currents. I'm not sure a typical gen-set is rated for that type of load, so you may need some additional k-factor rated transformers between your generator and miners.

You'd also need a good IT person.

Make sure you can get whatever permits are required. Probably need to get to know your gas/electrical/fire/building inspectors. I'd guess it helps to be off-grid, but you are at the mercy of their whim, and they are incentivized to be as restrictive as possible.

The trend has been that mining is going more and more industrial, and is basically a race to the bottom consolidating at the cheapest operating cost available. The higher your total cost per kwh the more risk. You mentioned hosing a bunch of your own S9s, so I'll assume you've done the calculation and are currently profitable running an S9, which would mean you think your cost will be under 3.5c /kwh. If that is true, then you're in the range (or close to it) that will likely stay profitable with moderately current miners no matter what the BTC price does. But it's just a matter of time before those S9s are no longer profitable with the new 30w/th miners coming to the market.

As for hosting, that's another headache. You'll probably need a lawyer to help with the contracts and/or when you get sued.....

Like Phil said, trust is an issue. It's fist a barrier to getting clients, and second to deal with issues after. The newer Bitmain miners have shown pretty high failure rates, your clients are going to have to take your word for it when you tell them it just died and wasn't because you let the temp get too high or your transformer blew up and fried it.

Don't think geothermal works very well with air directly, they normally use liquid and heat exchangers. And you might need 1000s (maybe 10s of 1000s...) of feet of pipe with liquid buried pretty deep in the desert to support the # of miners you can pack into a 900 sq ft shack.

I think you'd probably be better off using an evaporative cooling system, they work really well in dry places and can be made DIY pretty cheap.

edit: one estimate of pipe length per BTU I found after a little googling was 10 BTU per ft for dry soil. So for a closed system (with perfect insulation from the outside air), you'd need about 1000 feet per 3KW miner. Probably better to take the 105deg air from outside and pass it through the heat exchanger to bring it down to the temperature required and then just exhaust back outside, but even if it only required half the pipe, that's still a lot of digging for each miner...

Yeah, it seems like we might see a bunch of gear get turned back on. Could be a delay though because people are still freaking out about the virus.

I'm still waiting on a block at kano's pool. Estimated 47 days until the halving and we're down to 11PH. If the average diff till then is 15T then I have about 50% chance to get 1 or more blocks, 15% to get 2 or more, 3% to get 3 or more. Looking like my gamble to move over from viabtc is most likely going to be a looser. 2 blocks between now and the halving at the current pool hashrate would put me ahead though. If nothing else, at least I'm not sending my hash off to the evil empire....

I wouldn't be surprised if we get another drop next epoch as well. I think the media here (USA) is over-hyping the virus, but if half of what I'm reading is true then we will be dealing with the fallout of the virus for a long time.  Some studies have shown the R0 of the virus is between 2 and 3, which means it is way more contagious than the flu and all the measures going into place are only going to slow down the spread, not prevent it. Some experts are saying it is likely 60% of the world population will get infected before we're done with it.

China reporting 0 new cases ... people really believe that? I don't believe any information coming from the Chinese government is credible. Just read an article that says 21 million cell phone accounts were canceled in China in the last 3 months, and it suggests that way way way more (like 1000X more) people have died from the virus in China than they have reported. I really feel for the Chinese people, they are paying the largest price for their government's complete and total fuck-up of the handling of the virus.

Diff projection still heading down, the estimate is now sitting at -8.5%. At least it will be a good shakeout. For many miners that were on the cusp and barely profitably this was enough to motivate them to shut down now. Maybe many of those were trying to hold out until the halving and this got them out early. More BTC for those of us who are sticking it out for the long haul...

Numbers looking pretty bad, If price doesn't rebound we could see a big diff drop. Projected diff down around 5.5% at the moment.

On the bright side, we could see some sweet deals on s19s...

The price they are asking seems way too high for a miner that will only see a month of pre-halving operation (if they ship on time). The only way this price makes sense to me is if you expect BTC price to double or diff to halve before or very close after the halving. If that happens, then ROI with 5c/kWh power is about 250 days. If price and diff remain the same, then ROI with 5c/kWh power is 435 days, 500 days if they are late and you miss all pre-halving mining. Judging by how Bitmain has priced their miners in the past, I'm guessing they'll crush this price when they release their 5nm miners.