[Progress thread] Building out a crypto mine, step-by-step - AMA

[GlobalWealth]

Update:

Yesterday we had an interesting development inn the mine.

With crypto prices in such a dramatic decline, and especially ETH, we are rethinking the strategy to move some of our gpu rigs from the US to Estonia.

Not sure if I mentioned in earlier posts (too lazy to go back and reread my own posts...), but we have a mine operation running now in the US mining ETH.

My business partner and I invested with a friend of ours in an operation in the US early last year.

Now that ETH has been getting annihilated, the mine in the US is now significantly less profitable.

With energy costs around $.115/kwh in that area, the only ones making any money is the power company.

So we had a long call yesterday and it looks like we are going to ship hardware from the US to Estonia. The difference in power cost plus a much lower facilities cost will get it back into profitability. Of course shipping costs are going to eat into that, but we feel it's the right decision long term.

We aren't 100% decided on this, but we will finalize it by next week what we are going to do. Will keep you updated.

 

[GlobalWealth]

Update:

It's been awhile since my last update. It has been very busy getting this mine up and running. Imagine that, a new business that requires a lot of hard work..... go figure

We have shipped out all of your GPU mining rigs from the US going to Estonia. They are en route as I type and should arrive late next week or early the next week.

We have 19 rigs with 12 GPU's and 25 rigs with 8 GPU's. Total of 428 GPU's. We are running all RX580's.

This was a monumental task. My son handled this job exceptionally well. He went to the mine in the US to get trained on GPU mining, then disassemble, package and ship out the equipment.

In all, there were 4 full pallets with a total weight of nearly 3000lbs. Yeah, shipping was not cheap. Nor was it easy.

My advice for anyone attempting to ship a large shipment of electronics to Europe would be to get every single detail of every single component you are shipping. Label everything. Make sure it has the CE stickers on every component. Get the CE certificates for every component. Create very, very details invoice and packing lists. This was a huge mess and a huge hassle. Hoping it all pays off.

My son has been doing a lot of research on GPU mining and it seems right now, Monero is the most profitable option. At current rates, it is about 50% greater profitability than Ethereum with the same hardware.

Another interesting option is the Golem network. I am still researching that, but if you are GPU mining this may be an option for you. Basically you are "renting" your GPU processor for video editing work and get paid in Golem token. Whether you are a fan of Golem or a long term believer, who knows. But at least you can trade the coins for something else if profitability is good. Once we get the hardware in Estonia we will be testing it.

Between what we have running at home and sitting in customs as we speak, we have 12 Inno A9's and 12 Bitmain Z9 minis. We have 6 A9's running at the moment. I'm not plugging in anymore at home at risk of catching my house on fire (yes, that is a risk for all of you that think you can just plug in miners at home - the power consumption is way more than home wiring can handle).

This week I also bought 8 Bitmain L3+ LTC miners. I wasn't really interested in mining LTC, but I found a local guy who tried to get a crypto mining reseller business going but the coin prices tanked and he was desperate to unload equipment for cash.

I paid eur250 per L3+ including the PSU. They are all brand new in the box. Right now you can buy them from Bitmain direct for about usd300 with PSU plus about usd 100 for shipping, or about usd400.

So I save about usd100 per machine and have them in my hand today. LTC currently is not profitable to mine, but the difficulty right now is negative so based on projections using negative difficulty, they payoff for the hardware should be around 4-5 months.

Worst case scenario, I can at least reuse the PSU's, which cost about eur200 anyway. So it was a low risk investment.

The building is not yet finished and I'm getting pretty pissed at this point. The electricity is all done with the work on the transformer, new cables run to building and a 1000amp power panel. The last step is the electricians are running the cables to the mining area on Monday and Tuesday so I will be ready to plug in.

Unfortunately, the contractor has not yet finished the walls or the floor. I am going there again Monday and Tuesday to hold his hand and make sure he doesn't take 3 hour lunches it seems.

Regardless, it looks like we are plugging in miners in the property this week. Once the walls are up and the floor is ready, I will take some photos and post them so you all can see how the build out looks.

 

[GlobalWealth]

Update:

What a crazy week. Fighting with contractors, fighting with customs, fighting with electricians.....

We finally have our power all sorted. This was a monumental task.

I have attached a rough outline of the power map I created for this project. This is not a complete diagram, but if you are interested in mining, this should give you some ideas.

Basically we have a transformer with 630A available. We have stepped it down to 505A on our side of the building.

From there, we have high power cables into the building going to a 1000A power panel. This was intentionally overbuilt since everything will run 24/7 and we wanted to keep it all cool.

Inside the power panel, we have 30 - 3 phase 50A circuits. That means each circuit has 150A. If you do the math, that is 1500A. Since we only have 505A available, this again was overbuilt on purpose.

Right now we have installed 8 - 32A cables on 8 of the circuits. This means we have a max of 32A available on each cable, or 256A. This is basically 50% of our capacity at this point.

Most likely I will install another 8-12 - 32A cables so we are spreading the 505A load across 16-20 cables.

At the end of each of these cables is a 32A PDU (power distribution unit) with 4 - 16A outlets. The PDU has a 16A circuit breaker giving us further redundancy for any power spikes or overheating.

So this mean we have 8 cables with 8 PDU's with 4 outlets each, or total 32 outlets.

Since each PDU is limited to 32A, we can only plug in up to that limit.

This week we plugged in our Innosilicon A9's, Bitmain L3+'s, and Bitmain Z9 minis.

The A9's are drawing about 2.6A each. The L3+'s are drawing about 3.1A each.

With the Z9's, we learned something really cool. You can actually use one PSU (power supply unit) for 2 Z9's (we order one for one, but that is good since we need more PSU's anyway).

When I plugged in one Z9 it was drawing 1.3A. But when I plugged in 2 Z9's in one PSU, it was 2.1A, so we got a little bonus....lol.

So based on these numbers, we can now determine how and where to plug in which machines into which outlets.

So at each 32A PDU we will plug in 4 16A power strips. This means each power strip is limited to 8A (on average). So we can plug in 2-4 miners per power strip, average of 3.

So going backwards, we will have 3 miners on each power strip, 4 power strips per PDU, or 12 miners per circuit.

At this point I only have 8 cables, so I am limited to 96 miners based on these numbers.

The contractors are finishing the construction in the next 2 days and on Monday I am finally plugging in all of our hardware in the space.

Right now in our possession we have the following:

• 12 Innosilicon A9's
• 12 Bitmain Z9 minis
• 8 Bitmain L3+'s

From here, we will use profits for the first few months to rebuy hardware so we anticipate being at 50-60 ASIC miners by end of January.

 

[GlobalWealth]

One interesting note about the Z9 minis. From Bitmain they are preset with a 500mhz frequency as default. Right now we are seeing a hashrate of about 11ksol/s at this rate.

I have figure out how to adjust the frequency and have been playing with it a bit. I tested 650mhz and saw around 15ksol/s.

At 500mhz, the power consumption for 2 machines on one PSU was 2.1A.

At 650mhz, the power consumption for 2 machines on one PSU was 3.1A.

Right now I have all the Z9's set to 500mhz since we are limited on power capacity in the space where they are plugged in....(ask me how I know the limits...lol).

But Monday we will have the power capacity and I plan to test the machines at different frequencies to find the best sols/s vs. euro/kwh.

You can see the current dashboard of the Z9 mini here:

 

[GlobalWealth]

Quick update.

I just bought 80 Bitmain L3+'s for $85/e incl. PSU. Shipping cost is being finalized today or tomorrow but will run around $15/e.

At my power cost, my monthly profit per L3+ is about $50/month/machine (after power). So not huge money, but the payoff even at this rate is about 2 months (incl the extra cost of infrastructure: shelving, PDU's, power strips, cable, etc).

 

[GlobalWealth]

@GlobalWealth email me a sketch of your floor plan and the direction of your fans on the units and what other fans you have and we'll take a look at your airflow / cooling and see if we have any suggestions..... some baseline temp readings on and around would be helpful as well.....

If you look at the image with the huge yellow/black fan, there are 2 large windows that we removed from the frames. We put in metal screens and a filter on top so the fan pulls in outside air from those windows.

That huge fan blows directly toward the shelving with all the ASICS on it. The ASICS are situated with the intake fan side on the side of the huge fan. So the huge fan blows towards the miners, then the intake fan on the miners pulls air into the machine, and the exhaust fan blows out the other side.

If you see the image with the insulation board, we installed this board so that each miner has a hole cut out and the miner is pushed through the hole. This keeps the cooler air on the intake side, and the hot air on the other side of the heat wall.

We have "doors" on each side of the shelves so that the heat does not just go back around the shelving and intake hot air into the miners.

From there the hot air goes up and over the 2.5m wall.

There are some things we could (and likely will) improve.

We will most likely add a 2nd huge fan, and possibly a 3rd.

We are considering an exhaust fan on the roof to suck the hot air up and out.

Right now the temperature in the room (not the "hotbox" side) is around 25-30c. On hot days it can get to about 35c in the room.

The chip temps on the miners range from 64-71c. Board temps are about 10c lower.

We have almost all of the L3+'s overclocked right now and we are averaging around 600MH/s.

 

[biggeemac]

I've always had interest in doing this. I have two small rooms in the basement of one of my houses that I use for another business. This other business pays for the electricity at the house. I have thought about converting one or both of the rooms into a server room and running some crypto miners, but never really dug into the subject. I am an IT guy by day, so technically speaking, I should be able to figure things out. I know what my power costs, and its less than average according to what you say that the cost is in the US.

While I am not prepared to do the scale that you are, I would think running 10-20 miners in my basement may be possible. Thanks for doing this.

I guess my only question would at this point would be.....how do I figure out if this would even be worth my time and effort? And what sort of tasks would I need to do to transform a room into something that can reasonably handle a small operation like this?

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GuestUser4aMPs1]

I guess my only question would at this point would be.....how do I figure out if this would even be worth my time and effort?

Not an expert by any means, but there are tools out there that can calculate profitability based on your hardware/energy costs:

Mining Calculator Bitcoin, Ethereum, Litecoin, Dash and Monero

Or Google "Crypto Mining Profitability Calculator" and proceed to fall down the deepest rabbit hole in your life

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Scot]

Just musing here..

With all the hype around crypto right now, would making a mine startup company be a viable option?

Essentially, do what you’re doing now. Set up a warehouse, get 200 rigs going, show profitability in 6 months, then sell it off for a higher multiple.

Rinse and repeat.

Instead of having to manage these long term, worry about rig burn out, you can flip these every 6 months.

 

[GlobalWealth]

I've always had interest in doing this. I have two small rooms in the basement of one of my houses that I use for another business. This other business pays for the electricity at the house. I have thought about converting one or both of the rooms into a server room and running some crypto miners, but never really dug into the subject. I am an IT guy by day, so technically speaking, I should be able to figure things out. I know what my power costs, and its less than average according to what you say that the cost is in the US.

While I am not prepared to do the scale that you are, I would think running 10-20 miners in my basement may be possible. Thanks for doing this.

I guess my only question would at this point would be.....how do I figure out if this would even be worth my time and effort? And what sort of tasks would I need to do to transform a room into something that can reasonably handle a small operation like this?

For a small operation it is doable, but you should get an electrician to your house to make sure your wiring is capable of handling the power consumption. Most likely you will need to invest in some wiring upgrades and a new power panel with 50a breakers.

You need to consider heat dissipation also. If you are in a basement, you will need to channel the heat out. If your basement has an exterior window, you should be ok. 10-20 machines will create quite a bit of heat.

Sound will also be an issue. Miners are fairly loud. You don't want to listen to the humming of machines 24/7 while your trying to go to sleep at night. Maybe your basement is well insulated, but you may also need some sound proof insulation added.

Depending on what coin you want to mine, there are several online mining calculators out there. Just google search xyz coin mining calculator. You will need to get the power consumption figures from the mining rig (if you are buying asic miners, the manufactures will list the specs on their sites), the hash rate, and your $/kwh to calculate profitability.

 

[GlobalWealth]

For the first two machines you have up and running, are you actually getting near consistent 50Ksol/s as advertised?
I only ask because what I've read, some of the miners are reporting much lower hashrates than that. 30-35, some lower.

Great question and one of my biggest fears as well. They are averaging around 45k right now. Of course they've only been running a few days, but that is very consistent so far.

We've not done anything with them though. Right now they are just sitting on a table with the exhaust pointed out a window. We have not optimized anything yet. Once the construction is finished on the space, I imagine we will be able to run them a bit cooler and hopefully get a better hashrate.

It really seems like Innosilicon has been trying to flood the market as quickly as possible, thus offering a 2 for 1 buy-in which case it would seem that within a few months, the ability to profit would be considerably lower. This is my biggest concern as I just don't know how long this can be sustained re: profit calculations with more hands in the fire after Sept shipments.

Personally, I think this is a marketing gimmick and a way to just sell more equipment. The Inno A9 advertises a 50k hashrate and costs about 5k (after perpetual promo). The Bitmain Z9 (not the mini) advertises a 40k hashrate and costs about 3500. So the cost/hash is quite close (Bitmain is winning by a small margin, but the Inno model uses less power).

And of course they are selling a lot of them. You can go to their twitter feed and see the numbers they are selling (I don't recall right now). And yes, more machines will mean more competition. It can also mean more users and more growth also.

No one said this is the safest investment in the world....lol

Do you think that Zencash or Zcash might 'hard fork' thus causing the A9's to become obsolete overnite, like a few others?
These ASIC's seem like the way to go, but I'm just wondering if it is too late for a solo miner to profit or just break even.

I'm friends with one of the founders of ZenCash. I know that he has no issue with ASIC. The biggest issues with ASIC mining primarily comes from the GPU mining community. More of a sour grapes issue IMO.

He told me that they could do some upgrades to Zen that would require higher memory and could possibly brick some hardware that didn't come with enough memory to deal with the upgrades. But he has no agenda to brick ASIC like the Monero guys do.

With Zcash, the founder has publicly stated he has no plans to introduce any updates to their system that would make ASIC mining impossible. It is on the Zcash blog (google it).

These are some of the reasons we are initially focusing on ZenCash and Zcash. We do plan to mine other coins as we roll out the mine though.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[mtnman]

Very interesting thread. So without foreign tax benefits, and avg price per kilowatt in the US, this would essentially be a waste of time in the contiguous?

Not being cynical here, just ran some preliminary numbers and they don't back out like yours. I should also note, that I know nothing about mining anything other than gold.

I'm surprised no one is making earth sheltered greenhouses and lining these things up with geothermal cooling. It seems like it would be hard to control temperatures as a mining operation increases exponentially, without eating into the profits. Could get interesting though if you're able to harness the earth's constant temperature for free, independent of outside climate.

 

[garyfritz]

The major thing that has scared me away from mining is the way increasing network difficulty decreases profitability. I didn't have a good feel for how much of a problem that was, but @nitrousflame's charts give us some solid data to look at.

In the last year, difficulty has increased 3x. If we assume that trend continues, and the difficulty grows roughly steadily, that would say that difficulty increases -- and profits decrease -- by about 9% per month.

Here's a quick calculation assuming you buy one Z9 mini tomorrow (8/1) and net $400 / month after all power / fees / etc. The bottom line is the running profit.



So by risking $2000 plus whatever, it takes 7 months to break even, and after a year you have netted $870. Which is a 43% return and nothing to throw rocks at -- but it's not the $500 per month we were talking about. If you run the miners for another year, you net another $1000 or so, and by then the Z9 is only generating $50 per month.

If it turns out the difficulty increases 15% per month -- maybe due to the surge in difficulty that @million$$$smile predicted -- then after a year you clear a whopping $221 per Z9.

Does anybody see a flaw in this logic? If I'm right, it really seems to tilt the risk/reward calculation in a less-attractive direction. I'm not sure the modest reward is worth the risk of hard forks or other things totally out of your control.

 

[GlobalWealth]

Updates:

Last week we installed 8 more cables with 8 PDU's giving us more than enough capacity to plug in the remaining 25 V2's (12 GPU's per V2).

As of Sunday we now have all hardware at work.

The GPU rigs are currently mining ETH, which is not that profitable at the moment, but they were already configured for this and we are trying to get it all up and running as quickly as possible.

Over the this past week we ran more tests with XMR and it looks like the gross profit is not much different, but it is about a 20% power savings to mine XMR over ETH.

We will test a bit more but most likely get it switched over soon.

All of the A9's, Z9 minis and L3's are running fine and all still profitable.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GlobalWealth]

Loved following this thread @GlobalWealth. Hows the new project coming along with bulls coming back to play?

It's coming along nicely. I was in Germany 2 weeks ago buying a vanload of new asic hardware.

The mine has been profitable since day one. It took about a year or so to return the initial investment.

With each successive hardware purchase we have been (biased opinion) smart about our investments and have made sure to only purchase hardware when the return of investment is a maximum of 4 months.

We are nearly 2 years into the project and it is profitable and paying dividends to me and my partner monthly. It's not huge, but just another income stream.

Currently we are running the following hardware:

• 105 Bitmain L3+'s
• 12 Bitmain Z9 minis
• 14 Innosilicon A9's
• 10 Innosilicon A9+'s
• 63 Bitmain S9's

We are nearly at power capacity for our existing situation, but we are looking to add another 10 S9's and probably another 10 A9's. Once that is done, we will be at our limit.

All of the L3+'s have been overclocked and averaging 600mh/s.

I had previously overclocked the Z9 minis, but they aren't stable overclocked so much, but I have slightly tweaked them to get an average of 11k sol/s.

The A9's have all been overclocked to average about 60k sol/s.

I haven't figured out yet how to overclock the A9+'s so they remain at about 120k sol/s for now.

I have underclocked the S9's down to about 12.5 Th/s. I underclocked them to get more machines in the mine and found that the power savings to underclock was so big compared to the minimal decrease in Th/s. The result has been a higher Th/kwh. It has also allowed us to fit in more machines (higher net profit) and reduce the heat in the facility.

As you can see, it is a very small project but it is fun, interesting, VERY educational, profitable and aside from installation of new hardware takes very little time.

My business partner lives nearby and he goes to visit the facility 1-2 times per week (mostly because he is bored) and I am there only as needed, which is a couple of days every other month or so.

We have it all set up for remote access so I can do nearly everything from my laptop anywhere. The only things that require personal visits are if the modem or switches need to be physically rebooted or we blow a fuse (very, very rare), or we need to do a cleaning.

Additionally, we are building out nodes with some of our coin which generates an average of about 15% ROI (coin-coin roi).

Hope you find this interesting or helpful.

 

[garyfritz]

I've looked at this several times over the last 8-10 years, but I always concluded the cost of entry, the razor-thin profit margins, and the short lifespan of miners before they became too weak/slow to continue producing good results, all combined to make it not viable. Apparently I've been wrong because smart people are still doing it. Looking forward to your updates, GW!

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GlobalWealth]

What critical piece of information am I missing?

LOL. That is a BTC miner, not ETH. The payoff on that machine is around 2 years. Not a great deal IMO.

 

[GlobalWealth]

In Estonia, all company information is publicly available.

BTW, reason 12,365 the Estonian e-residency program is complete bullshit.

Absolutely zero privacy.

 

[GlobalWealth]

That spreadsheet looks mighty attractive. $6000 / mo out within a few months!? Wow. That's with 12 Z9's, so you'd have an initial outlay of $24k plus any additional infrastructure.

The key word here is additional infrastructure.

Between construction costs and electrical infrastructure, I have invested about eur42,000.

This is the problem with small/midsize mining. The infrastructure costs require that you do a reasonable size operation to overcome those costs.

So basically, you can plug in a couple of machines in your basement and hope your wiring doesn't fry and catch your house on fire (actually a real risk), or invest several hundred thousand or more. There's no middle ground.

But the thing that's always stopped me is... don't these miners have a very limited lifespan? Doesn't the problem difficulty increase so fast that a miner is basically obsolete within a few months? How much of a productive lifespan do you get before the miner chews up more power than it's worth? I got the impression the miner value tended to go to zero just about the time it paid itself off, so you're always on a hamster-wheel of trying to squeeze out profits before the miner loses its punch.

For sure there is a limited lifespan with miners. Based on everything I've read, that's about 2 years. The difficulty does increase and new hardware comes out that is much faster making your old slugs obsolete.

We are anticipating this for sure. We know a certain amount of profits will need to be rolled forward to keep the lights on with new hardware. No doubt.

But I don't see any ongoing miner-upgrade costs in your spreadsheet. So you're planning to run the miners for months, producing a more or less constant income??

This spreadsheet is only until the end of 2019. I intentionally did it like this because that is about how long it will take us to reach capacity with our existing power limitations.

At some point, my business partner and I will need to decide if we are going to build a 2MW transformer, or just keep slowly upgrading hardware to maintain the current state.

I don't know what we are going to do at that point yet.

But I can tell you this. I am 44 years old, my business partner is 50. If we get anywhere close to our model of profitability, I may find out I am too lazy to grow the mine any larger.

Or maybe I get really greedy....lol

 

[GlobalWealth]

Also, do you have a preferred mining pool for these machines?

Sorry, I missed this question earlier.

I'm using zen.zhash.pro. This was the pool recommended to me by the ZenCash guys.

 

[GlobalWealth]

Quick update.

For those of you interested in predicting future profitability based on the difficulty increasing, you might find this site useful (at least for the coins they cover).

My Crypto Buddy

You can plug in the basic details like power usage, hashrates, power cost, and pool costs. But you can also plug in hardware costs, difficulty increase, and selling coins to predict future profits.

 

[GlobalWealth]

Some lessons learned.

Familiarize yourself with the site rapidtables.com.

You will use this often to calculate the amp draw of your hardware. For example, an Innosilicon A9 overclocked needs about 1.7amps using a .8 power factor.

A Bitmain L3+ overclocked at 462mhz needs about 2.5amps using a .8 power factor.

Our 8 GPU mining rigs need about 3.5amps and the 12 GPU rigs need about 4.4amps using a .8 power factor.

The Bitmain Z9 minis (not yet tested, but based on stated wattage) need about 1.3amps using a .8 power factor.

Electricity infrastructure is BY FAR the biggest factor to work out. Do not underestimate this.

Do not assume you can just plug in miners and you make money.

For example;

At this point, I have 504amps available. My power panel is rated for 1000amps. I have 30, 3-phase 50amp circuits. Monday I am running 8, 32amp cables into 8, 32amp PDU's. Each 32amp PDU has 4 16amp plugs.

Later I will run another 12+ 32am cables with 32amp PDU's.

If you do that math, 4, 16amp plugs on each 32amp PDU is 64 amps. That means you can only use 32amps so each of your 4 plugs cannot exceed 32amps combines, not the 64amps it is rated for. So you need to do a lot of math to determine what kind and how many miners you can plug into each plug. We will be using high voltage power strips plugged into the 32amp PDU, but most likely will only be able to plug in 3-4 miners per power strip depending on what kind of miner it is. We will also need to do some mixing and matching in order to balance out the power usasge.

If you have no idea what that means, I suggest some studying.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GlobalWealth]

Thanks for the update! Have been thinking of you lately, what with all this crypto blood on the streets!

Blood in the streets is the scariest time to invest in any asset class. It is also usually the best time to make money.

I made a killing in 2008-10 in the stock market when no one wanted to touch stocks.



Sent from my VTR-L29 using Tapatalk

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[ZCP]

@GlobalWealth my company is an engineering consulting firm. we do airflow studies and consulting for industry. send info, dimensions, specifics and i can probably hook you up for free (a few drinks). then you would benefit from math and science!

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GlobalWealth]

Could you share why if that's not private?

Partnership disputes.

I had a partner and it was a clear situation of "too many cooks in the kitchen".

He was more involved locally so he bought me out.

It worked out great actually for both. The business was growing as we were taking on clients so he will continue what we started and I cashed out nicely.

I'm in the process of launching a new company in my current space so the timing was good to allow me to focus.

I'm also seriously considering to launch a private crypto fund that would pay monthly dividends.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[MTF]

Good luck! I have no idea about crypto mining but will follow this thread anyway.

 

[GlobalWealth]

Hah! A word of warning to anyone looking at this.

I started with 1 gpu out of curiosity, and now have 30

We also started with gpu mining with another partner. Now have a couple thousand gpu's humming and heating a warehouse.

 

[GPM]

So these guys wanted to recoup their investment, but I could have replicated their operation at just a bit more than half what they invested.

That is exactly what made me build my second GPU rig. I was looking for used stuff and everyone wanted like $4,000-7,000 for some hardware that had been used for 6 months. I got to researching the costs to building a brand new one with better hardware and it was cheaper than most peoples used stuff. And it just kept growing from there...

Are you doing anything fancy for your cooling on this, or do you just route all the air in channels in and then out of your mining areas?

Are you also worried about security at all on all these fragile machines?

You say you have investors on this project correct? How does that look like, are you cashing out any coins to pay them back at all, or is everyone on board with holding until target prices are reached?

 

[CareCPA]

LOL. That is a BTC miner, not ETH. The payoff on that machine is around 2 years. Not a great deal IMO.

That makes more sense than the billion percent returns it was forecasting.

So do you find the income amounts on this page to be accurate: Realtime mining hardware profitability | ASIC Miner Value
It projects ~$100 per day at current prices for the A9 ZMaster.

 

[Roli]

Step one:

Choosing a location.

We did a lot of research on power costs, power infrastructure, crypto regulations, ease of access, cost of real estate and more.

First of all, we have excluded the US as a viable option for any reasonable sized mining operation.

Regulations and the future of additional scrutiny are the main reasons. However, power cost was also a significant concern.

The average cost of power in the US is around $.105/kwh. While not the highest in the world, even a 10% reduction in power cost has a significant impact on the bottom line when you start discussing any amount of scale.

Of course if I'm just going to plug in 1-2 miners in my garage, it's not an issue. But we are planning to have 100-200 miners running within 6-9 months.

Of course there are low cost power options in the US, but typically that requires a HUGE upfront investment in infrastructure in order to control your own power source and get to a reasonable cost.

I've been to a few conferences over the past year or so and met guys getting .05-.07/kwh in the US, but that required signing contracts for 10mwh+ or buying their own hydro stations. I didn't have access to enough funds to get to that level.

Beyond power cost, heat management is a huge concern. We had an option to build a facility in Georgia (US) for a very low upfront cost, but the power cost was .113/kwh plus the heat in GA is so high, we would spend an enormous amount of money (and reducing our kwh capacity) for air conditioning to get to a reasonable operating temperature.

We have looked at several other options including Russia and Georgia (the central Asian country), but opted out of those for various reasons.

We are keeping Georgia (country) as an option since we have very good connections there.

Ultimately, we settled on building the facility in Estonia.

I will discuss this in further detail in a later post.

Great thread. I think I can already guess why you chose Estonia, I've been thinking of becoming an e-citizen of them for awhile now.

In my mind they are miles ahead and it shouldn't be too long before they are the place to go for all things tech.

I'm guessing they are giving you huge incentives to set up there? Plus of course their broadband coverage is second to none.