DataTank Mining: 1.2MW 3M Novec Immersion Cooled 2PH Mining Container

[mc_lovin]

can we buy entire containers ready to plug in?

[Silverspoon]

^Yes, but you need one of those 220 dryer outlets.  Got one?

[mc_lovin]

^Yes, but you need one of those 220 dryer outlets.  Got one?

Maybe I can take the diesel V8 and convert it to run on vegetable oil and then drive around the fast food joints and get the free grease all day for fuel..  MMmm..  How many alternators do I need to rig up for that?

[jeycreative]

How would DataTank Mining's interests be aligned with shareholders'?  As far as I can tell, once the container has been sold, DTM just makes a tiny cut of electricity bills, so your only incentive is to keep the container operating?

The cost of operation and building of DataTank infrastructure is passed through, there is no profit involved.

DataTank Mining's profit comes from deploying 20% of _identical_ hardware alongside the public capacity. Hence, we only earn money if we are successful. The 20% fee on operating cost is used to fund the day-to-day operation, including on-site staff providing various services such as ongoing hardware re-deployments, maintenance and security.

Thanks.  For clarity, does the IPO also fund your 20% deployment of identical hardware, or will you be coughing up 2 BTC of your own for every 10BTC invested?

DTMB holders can essentially convert to DTMA-equivalent with the at-cost purchase of mining hardware right?  ie, for 0.0165 BTC / unit at current prices. 

How will you handle DTMB holders who want different things?  Eg, some might want you to help negotiate to franchise their capacity.  Others might want to handle their own franchising, while yet others may want to self-mine.  Seems like it would be messy to communicate with individual shareholders going in different directions.  If the shares are not fungible, there will be more complications when trading shares and issuing dividends on havelock.

+1 on that. how we will get in touch with you guys ? have you already a website designed to handle change requests for B stakeholders ? And for the A Team, can people choose where they are mining ? or are you creating a pool ?

[tinyfox266]

electricity cost / mining revenue < 10% nowadays, why a person will rent the capacity by sacrificing 20% revenues?

[antirack]

electricity cost / mining revenue < 10% nowadays, why a person will rent the capacity by sacrificing 20% revenues?

I am trying to understand this.

Do you mean 'income' when you are saying revenue?

You are saying that net income after paying electricity cost is only 10%?
Wouldn't this depend on miner and electricity cost?

Electricity cost TerraMiner IV = $5.28 per day at 2.2kW at $0.10/kWh
Mining Income TerraMiner IV = $35.66 with 1.6TH

And where does the "sacrificing 20% revenues" come from?

[tinyfox266]

Do you mean 'income' when you are saying revenue?

yes

Electricity cost TerraMiner IV = $5.28 per day at 2.2kW at $0.10/kWh
Mining Income TerraMiner IV = $35.66 with 1.6TH

$5.28/$35.66 = 14.8% < 20%, sacrifice 5% still.

[cooldgamer]

This is looking good, can't wait to get my hands on some DTMB shares

[open_sauce]

Does the IPO start at midnight in Havelock's timezone on the 24th? Can't find any mention of time.

[DataTankMining]

Does the IPO also fund your 20% deployment of identical hardware?

That is correct. The 20% also function as safety buffer in case of price fluctuations.

DataTank: Sources and produces ASIC boards in-house without third parties involved for significantly more than 20% cheaper (relies on experience producing past ASIC boards and direct connections to designers/manufacturers). There are no profits involved through the complete chain and we are on-site for manufacturing, and on-site for deployment. Case in point. We produce hardware at today's prices for <$438/T (<$0.44/G). This compares to the price of other hardware available.

Traditional: Buys ASICs, components and PCB from third parties. Relies on manufacturing for third party. Needs to add mechanical components such as case, fans, heatsinks, screws, nuts, switches, cables.  Needs box and packaging. Needs logistics. Needs profit margin to make money. Needs to sell with customer support, billing, needs to cover NRE expenses. This total cost will be significantly more than 20% margin. In addition, when the end-user receives hardware, the costs continue. Power supplies, installation, deployment, overpriced data center rent, etc
These are some of the problems that DTM solves.

[antirack]

Do you mean 'income' when you are saying revenue?

yes

Electricity cost TerraMiner IV = $5.28 per day at 2.2kW at $0.10/kWh
Mining Income TerraMiner IV = $35.66 with 1.6TH

$5.28/$35.66 = 14.8% < 20%, sacrifice 5% still.

I am still not following. Nobody is charging 20% from your mining income and nobody is charging 5%.

Instead of the TerraMiner IV example (around 1.6TH) I should have used the real DTM numbers:

electricity cost for 1.66TH at DTM = $0.6/day ($18/month)
20% fee on the electricity cost for operating the cluster = $0.12/day

why a person will rent the capacity by sacrificing 20% revenues?

I certainly hope that 30 * $0.12 = $3.6/month is far away from 20% mining income on 1.66TH for a while.

If it isn't then even DTM's efficient system won't help (and certainly not paying MUCH MORE for a data center or home mining where you'd have to fork out the annual mining income or more just for the shipping cost).

[DataTankMining]

how we will get in touch with you guys ? have you already a website designed to handle change requests for B stakeholders ?

DataTank Mining will have a control panel for unit holders to control their capacity.

In addition, there is a NOC (network operations center) staffed with engineers that will respond 24/7 to requests in addition to securing networks. This is part of the security protocol.

There are currently discussions with Havelocks and considerations for other (also open) exchanges to integrate an interface (API).

The easiest and solution already in place is moving units between the unit holder and DTM's account manually, but this introduces a lag due to manual interaction. We are confident that this can be automated, despite the the new twists that the systems introduces.

or are you creating a pool ?

A pool and a digital market place is under consideration, with third parties that are established with a proven track record in this space.

can people choose where they are mining ?

Yes. But DTM can only offer additional benefits (digital market place, ie. additional income via selling hash rate to cloud miners or cloud operators) when miners are pointing at DTM systems, not at external pools.

DTMB holders can essentially convert to DTMA-equivalent with the at-cost purchase of mining hardware right?

That is correct, via control panel or manual interaction.

How will you handle DTMB holders who want different things?  Eg, some might want you to help negotiate to franchise their capacity.  Others might want to handle their own franchising, while yet others may want to self-mine.  Seems like it would be messy to communicate with individual shareholders going in different directions.

This is a matter of proper planning and implementation. The mining market will definitely not stand still, so this will result in a ongoing effort.  Control panel, on-site technician and network operators will be prepared to handle requests and keep track.

[horcoff]

Does the IPO also fund your 20% deployment of identical hardware?

That is correct. The 20% also function as safety buffer in case of price fluctuations.

DataTank: Sources and produces ASIC boards in-house without third parties involved for significantly more than 20% cheaper (relies on experience producing past ASIC boards and direct connections to designers/manufacturers). There are no profits involved through the complete chain and we are on-site for manufacturing, and on-site for deployment. Case in point. We produce hardware at today's prices for <$438/T (<$0.44/G). This compares to the price of other hardware available.

Traditional: Buys ASICs, components and PCB from third parties. Relies on manufacturing for third party. Needs to add mechanical components such as case, fans, heatsinks, screws, nuts, switches, cables.  Needs box and packaging. Needs logistics. Needs profit margin to make money. Needs to sell with customer support, billing, needs to cover NRE expenses. This total cost will be significantly more than 20% margin. In addition, when the end-user receives hardware, the costs continue. Power supplies, installation, deployment, overpriced data center rent, etc
These are some of the problems that DTM solves.

As I can read, Your container with power supplies without chips costs for You 0.38 $/GH (or 0.633 $/W with Your assumption of 2 PH per 1.2 MW container). In Your reports were instalations with as low as 167$/kW of capital costs.

1 share of DTMB is going to IPO with a price of $7.25 per 10w (0.0115 btc per share, page 9 in Prospectus) or $0.725 $/W (15% more than costs).

Worse thing with DTMA: You claimed "<$820/T (<$0.82/G) total cluster with deployed DataTank system on-site" of costs or 1.37 $/W. At the same time You are going to sell 1 DTMA share per 0.0280 btc or 1.76 $/W (29% more than costs)

You try to earn 15-29% on IPO, with a history of making some immersion cooling systems - it is fair. All clients funds are in Your hands, so You will be very liquid.

You are trying to sell 1 GH of mining power in low cost of electricity location in the end of 2014 with price of 0.62 $/GH (DTMA case, pure mining boards without container and without power supplies) when I can buy Antminer S2 or Rockminer hashing now for 1,5 $/GH. The difference will be mined in 2 months.

In DTMB case You are going to sell 1 kW in container per $725 in the end of 2014 and are going to rent them (if fully deployed) for $60 per month (with $18 of operational cost). It's 17 months to get invested money back. It is for long term investors considering the average period of 4-5 months in case of buying miners (which usually could work a year).

PLEASE, make things clear: (1) if we give You say until June, 30 the sum of $870 000 as You ask for 1 container, when it will be ready and (2) when ASICminer chips could start mining in container if we give You
$2 112 000 until June, 30 as You ask for one container fully deployed with chips?

PS. The faster You earn people 50% - the quicker they will give You 10x more money.

[jimmothy]

As I can read, Your container without chips on boards and power supplies costs for You 0.28 $/GH (or 0.168 $/W with Your assumption of 2 PH per 1.2 MW container).

It's $0.46/W.

1 share of DTMB is going to IPO with a price of $7.25 per 10w (0.0115 btc per share, page 9 in Prospectus) or $0,725 $/W (4.3x more than costs).

They are asking $6.80 per share not $7.25.. And it's 7% markup not 400%.

Better thing with DTMA: You claimed "<$820/T (<$0.82/G) total cluster with deployed DataTank system on-site" of costs or 0.49 $/W. At the same time You are going to sell 1 DTMA share per 0.0280 btc or 1,76 $/W (3.6x more than costs)

It costs them $1.36/W and they are asking $1.65/W.

You try to earn 3.6-4.3 times on IPO, without even previously done Your own container with Your own money. In such a case profit margin from IPO cannot be more than 50-100% depending on business model.

They are not asking for nearly that much and they have successfully built immersion datacenters before.

MOREOVER You are trying to sell 1 GH of mining power in low cost of electricity location in the end of 2014 with price of 1,725 $/GH (DTMA case, mining boards and power supplies without container) when I can now buy Antminer A2 or rockminer hashing now for that price.

They are not selling individual GH nor for that price. Each share represents about 17GH and costs about $17. Also there is no such thing as an antminer A2 and there is also no miner for less than $1/gh (or not that I am aware of). Also I am 99% sure you do not pay $0.025/kwh so there's that.

Moreover Your profile here without previous history.

Look up allied control. Plenty of history.

[Collider]

They are not selling individual GH nor for that price. Each share represents about 17GH and costs about $17. Also there is no such thing as an antminer A2 and there is also no miner for less than $1/gh (or not that I am aware of). Also I am 99% sure you do not pay $0.025/kwh so there's that.

Electricity cost in WA state for Datacenter use goes for 3-4ct/kWh, so I am quite confident that described rates can be achieved.

In any case, it is much lower than traditional hosting and is also more power efficient (cooling).

[jeycreative]

Thanks for your answers !!
I'm satisfied.

[horcoff]

Edited my post, please edit Yours if quoted.

[Collider]

PLEASE, make things clear: (1) if we give You say until June, 30 the sum of $870 000 as You ask for 1 container, when it will be ready and (2) when ASICminer chips could start mining in container if we give You
$2 112 000 until June, 30 as You ask for one container fully deployed with chips?

This is in the prospectus. It takes between 3-6 months for full deployment. The longer it takes the cheaper hardware per GH gets (OR you receive more efficient hardware, perhaps from spondoolies, or another GEN....)

As hardware will only be bought (produced) once the date for DataTank deployment should be fairly firm, the cost and efficiency can only be determined then.

However, for a best prediction of cost, DTMA holders are asked to pay a fee per kW now, which shouldn´t change as much as the price per GH over time, when more power efficient equipment becomes available.


As i hopefully described comprehensively, it is less relevant when the actual Tank will be ready, as the hardware prices at the time of deployment will be used for calculating the cost per GH.

[Ozziecoin]

PLEASE, make things clear: (1) if we give You say until June, 30 the sum of $870 000 as You ask for 1 container, when it will be ready and (2) when ASICminer chips could start mining in container if we give You
$2 112 000 until June, 30 as You ask for one container fully deployed with chips?

This is in the prospectus. It takes between 3-6 months for full deployment. The longer it takes the cheaper hardware per GH gets (OR you receive more efficient hardware, perhaps from spondoolies, or another GEN....)

As hardware will only be bought (produced) once the date for DataTank deployment should be fairly firm, the cost and efficiency can only be determined then.

However, for a best prediction of cost, DTMA holders are asked to pay a fee per kW now, which shouldn´t change as much as the price per GH over time, when more power efficient equipment becomes available.


As i hopefully described comprehensively, it is less relevant when the actual Tank will be ready, as the hardware prices at the time of deployment will be used for calculating the cost per GH.

I thought we were paying $/W and not $/GH? Can anyone confirm?

[DataTankMining]

Previous posters are correct.

DataTank Mining is not selling or purchasing any hardware (as in hashrate, ASICS, etc) today, and prices are merely a representation/reference to make a business case. It is at the core of the business strategy _not_ to engage in such practice and bet on hardware. Hardware will be purchased in time for deployment, no earlier than that, at the price and performance that is best suitable to serve its purpose.

Hardware acquisition needs to happen at a certain point in time, no doubt. Because numbers cannot be pulled out of thin air, we are using today's known data for reference calculations. However, it can be safely assumed that the hardware will be cheaper (more competition) and that the same money can buy more hashrate (advancement of technology) in the future. Both cases are accounted for in the insurance sections: unit holders receive excess funds paid back via dividends or receive more capacity/higher hashrate in form of extra units if we are over-funded.

The reason why today's known hardware cost (large downwards trend) and production cost (very little change) are used, instead of guesstimating or making up numbers is because it's bad business practice to base a business plan on speculation.

Keep in mind that we run our own production line. We have produced ASIC hardware for third parties before, not only for immersion but also for heat sink users worldwide. Prior to that, early 2012, we have built and produced Spartan 6 based FPGA boards and there are surely a few forum users who could attest to that (info via PM if needed).

-- // --

Having said that, there will be a progress update within today, of which one particular point addresses exactly the risk of hardware acquisition. We have some good news in this regard but discussions are still ongoing.