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

[antirack]

I simply give up. You are not a friend of this technology and you believe I am misleading people. I accept your point of view of course.

Can you please confirm or deny that a business plan with a giant heat sink (and apparently the invention of a complete new power distribution system) has been interrupted? In other words, do you believe what you are saying because you are fascinated by technology, or is it because you are trying to create your own efficient and competing solution to what DataTank is? The latter would explain a lot.

DC Power:

I am all for new solutions that are cheaper (evidently by the use of immersion), but I need to build mines with existing technology. Highly efficient 12V power supplies at less than $0.10/W are part of this. DataTank also doesn't mind using 48V or even 480VDC. Anything fits in them. But currently not too economical, without reinventing the wheel, but trust me when I say there are tests being down and there is progress in this direction  Will it be ready in a few months for deployment? Probably not... there are also some ASIC/board manufacturers looking into this direction and a whole industry is trying to create more efficient DC power systems.

Space requirement:

1 container can hold the equivalent of 576 SP30. It has the foot print of exactly 1 container. With everything. I do honestly not understand what you are trying to say and why you think it would require the spac3e of 6-9 container units etc.

[jimmothy]

2112, How much money do you think it would cost to build a cheap as possible air cooled 1MW DC? (including the cost of the warehouse, heatsinks, fans, AC, psus, basically everything but the asics themselves)

[antirack]

2112, How much money do you think it would cost to build a cheap as possible air cooled 1MW DC? (including the cost of the warehouse, heatsinks, fans, AC, psus, basically everything but the asics themselves)

He is promoting a giant water cooling system I believe, not air cooling. But I am sure he is also knows how to build cheap large scale air cooled facilities and is teaching Dave Carlson of MegaBigPower fame as we speak (just kidding).

We will soon know his costs when he is finished building his first MW!

Using Hammer or Rockerbox chips I assume, and kilovolt voltage DC power supplies.

[2112]

I simply give up. You are not a friend of this technology and you believe I am misleading people. I accept your point of view of course.

Can you please confirm or deny that a business plan with a giant heat sink (and apparently the invention of a complete new power distribution system) has been interrupted? In other words, do you believe what you are saying because you are fascinated by technology, or is it because you are trying to create your own efficient and competing solution to what DataTank is? The latter would explain a lot.

DC Power:

I am all for new solutions that are cheaper (evidently by the use of immersion), but I need to build mines with existing technology. Highly efficient 12V power supplies at less than $0.10/W are part of this. DataTank also doesn't mind using 48V or even 480VDC. Anything fits in them. But currently not too economical, without reinventing the wheel, but trust me when I say there are tests being down and there is progress in this direction  Will it be ready in a few months for deployment? Probably not... there are also some ASIC/board manufacturers looking into this direction and a whole industry is trying to create more efficient DC power systems.

Space requirement:

1 container can hold the equivalent of 576 SP30. It has the foot print of exactly 1 container. With everything. I do honestly not understand what you are trying to say and why you think it would require the spac3e of 6-9 container units etc.

You are free to think about me what you like. I know who I am, and that I'm not affiliated with any pro-Bitcoin or anti-Bitcoin organization.

On the other hand from you I get a strong Amway/Quixtar vibe. Except that in your case you are a 3M-captive sales-person or sales-engineer.

Until you started posting in various threads "3M Novec will let you get rid of heathsinks, fans, water, cables, etc..." I thought that Novec is a solvent. I've seen Novec name multiple times on the empty solvent containers, next to trichloroethylene, tetrachloroethylene, isopropyl alcohol, etc. in a nearby trash storage area for a medical facility. There were all kinds of generic names there, but two proprietary brands caught my eye: 3M Novec and Solvay Galden.

Before writing this post I did both Google and local search on this forum for Galden. This mention is going to be the first.

I've met many honest consulting engineers and honest investment advisors. All of them make a strong ethical point for doing a proper competitive analysis of the field before they give an advice. How come you as a "immersion cooling specialist" never mentioned a leading competitor to 3M? Because you are strongly encumbered by your affiliation. And this is a serious red flag, when you never mention potential competitors or risks. What else are you hiding in your marketing materials?

It is difficult to get a man to understand something, when his salary depends upon his not understanding it!

[2112]

2112, How much money do you think it would cost to build a cheap as possible air cooled 1MW DC? (including the cost of the warehouse, heatsinks, fans, AC, psus, basically everything but the asics themselves)

There are multiple problems with your question. It is seriously slanted.

1) DC - I believe you meant datacenter. When I write DC it always means "Direct Current" and I always spell out  the whole word datacenter.

To repeat myself from just previous page: Bitcoin mine is neither a datacenter nor a supercomputer. It is a mass quantity of unusual chips with very rare property: they are all about gazing at their own navel and thumb twiddling, they do not communicate with each other and barely communicate with the external environment.

2) AC - I believe you meant air conditioning. When I write AC it always means "Alternating Current" and I would always spell that out.

Again I repeat myself: air conditioning is not required. Bitcoin mine is basically a giant distributed semiconductor resistor, unusual with the requirement that the voltage it dissipates has to be less than 1V with somewhat specific tolerances. For such a resistor air conditioning is not required, there are many other ways to evacuate the waste heat of computation/thumb twiddling.

So I'm basically left with the following possibilities:

A) you are trying to troll me
B) you haven't read and understood what I wrote
C) you are trying to use a strawman argument

Which one of A,B,C applies here?

[jimmothy]

2112, How much money do you think it would cost to build a cheap as possible air cooled 1MW DC? (including the cost of the warehouse, heatsinks, fans, AC, psus, basically everything but the asics themselves)

There are multiple problems with your question. It is seriously slanted.

1) DC - I believe you meant datacenter. When I write DC it always means "Direct Current" and I always spell out  the whole word datacenter.

To repeat myself from just previous page: Bitcoin mine is neither a datacenter nor a supercomputer. It is a mass quantity of unusual chips with very rare property: they are all about gazing at their own navel and thumb twiddling, they do not communicate with each other and barely communicate with the external environment.

2) AC - I believe you meant air conditioning. When I write AC it always means "Alternating Current" and I would always spell that out.

You understood both those abbreviations perfectly fine so I would say I am not misusing them. And a bitcoin mine is a datacenter. Not sure what is even debatable about that.. Regardless of the semantics you completely avoided the question.

Again I repeat myself: air conditioning is not required. Bitcoin mine is basically a giant distributed semiconductor resistor, unusual with the requirement that the voltage it dissipates has to be less than 1V with somewhat specific tolerances. For such a resistor air conditioning is not required, there are many other ways to evacuate the waste heat of computation/thumb twiddling.

Air conditioning is not required? So you know something that everyone with a datacenter in warm climates doesn't?

So I'm basically left with the following possibilities:

A) you are trying to troll me
B) you haven't read and understood what I wrote
C) you are trying to use a strawman argument

Which one of A,B,C applies here?

It seems like A and C perfectly apply to you. As for B, I believe you understood the question.

Again how much do you think it would cost to build a 1MW "bitcoin mine"? (excluding the asics)

[2112]

You understood both those abbreviations perfectly fine so I would say I am not misusing them. And a bitcoin mine is a datacenter. Not sure what is even debatable about that.. Regardless of the semantics you completely avoided the question.

Air conditioning is not required? So you know something that everyone with a datacenter in warm climates doesn't?

It seems like A and C perfectly apply to you. As for B, I believe you understood the question.

Again how much do you think it would cost to build a 1MW "bitcoin mine"? (excluding the asics)

OK, so you decided on sophomoric trolling.

However the last question is sensible (without the parenthetical qualifier).

Again how much do you think it would cost to build a 1MW "bitcoin mine"?

My answer will be to the general readers of this thread, not to jimmothy in particular.

That question is interesting. I assume that 1MW was just a shorthand for "industrial size" sold in 1MW increments. With this type of investment you can afford to hire the services of two engineers:

1) ASIC engineer with knowledge of mixed-signal and/or analog IC design including the IC packaging. It doesn't take a rocket surgeon to understand that plastic QFNs and thousand pad BGAs aren't the proper solution. I've actually given this advice to friedcat long time ago, he replied and then quickly deleted his reply.

https://bitcointalk.org/index.php?topic=91173.msg1062969#msg1062969

2) chemical process engineer with knowledge of thermodynamics-intensive processes and basic understanding of electromagnetic field theory or electrical engineering. He'll help you design the proper heat exchange system while paying attention to the electrical power&signal requirements of the chips.

Just let the two of them work for couple of days/weeks and they will come with a truly competitive Bitcoin mining installation.

[jimmothy]

So basically you are saying immersion cooling is too expensive but you have no idea what it costs to build a 1MW farm?

[horcoff]

So basically you are saying immersion cooling is too expensive but you have no idea what it costs to build a 1MW farm?

)))

I guess You are discussing with HR director )

[antirack]

you are a 3M-captive sales-person or sales-engineer.
..,
How come you as a "immersion cooling specialist" never mentioned a leading competitor to 3M? Because you are strongly encumbered by your affiliation.

Personal accusations which are not very helpful.

I mentioned above in reply to one of your posts that DataTank works with any fluid. The web site also states from day one the following:

DataTank Mining will also use different types of fluids and brands of fluids to further reduce costs where possible.

http://www.datatank-mining.com/faqs/wont-2-phase-immersion-cooling-cost-more/

Feel  free to compile a list with competing products, compare physical properties, cost, availability and all that. You will do me a big favor, as I will be able to compare that with our own findings and do some test boiling if we haven't done it yet. We've been doing this for a few years now, IRL you know

[Collider]

I don´t think he was misleading anyone when stating "no more heatsinks and fans required" as he was obviously talking about the mining units themselves, not the whole solution.

I think you are nitpicking answers, taking them out of context and then adding some arbitrary rambling to it so somebody might think you have a valid point.

[Ozziecoin]

So basically you are saying immersion cooling is too expensive but you have no idea what it costs to build a 1MW farm?

)))

I guess You are discussing with HR director )

He got me a bit worried... Lol 

Antirack, to dispel all doubts can you categorically say all reasonably known costs have been included in your estimates?

[2112]

So basically you are saying immersion cooling is too expensive but you have no idea what it costs to build a 1MW farm?

Actually let me write a short summary:

Datatank advocates two-stage Novec+water cooling system that is two-phase immersive in its first stage. The successfully deployed pilot installations used either chips unsuitably packaged for high-power application (Xilinx Spartan) or used under-engineered ASIC chips&boards from a vendor who already admitted to not being able to hire suitable engineering talent. All Datatank systems already/will contain a full power water cooling system (either open-loop or closed-loop) that works as a second stage of their design. In my estimation a large scale dedicated Bitcoin miner will realize savings in both capex and opex by utilizing a single-stage liquid cooling system that is essentially equivalent to just a second stage of the Datatank system. Those savings are contingent on being able to hire suitable engineering talent with expertise in high-power and mixed-signal installations.

In my estimation the only way where investing in Datatank securities and/or equipment makes sense is when the Bitcoin mining is just a temporary stage in a long term strategy to compete in the compact/nomadic data center market currently served by e.g. SGI.

[antirack]

Antirack, to dispel all doubts can you categorically say all reasonably known costs have been included in your estimates?

Yes, of course. All costs are included. Including every last screw and nut and a concrete slab for it to sit on. Contrary to 2112's belief, we are not beginners

[Ozziecoin]

Antirack, to dispel all doubts can you categorically say all reasonably known costs have been included in your estimates?

Yes, of course. All costs are included. Including every last screw and nut and a concrete slab for it to sit on. Contrary to 2112's belief, we are not beginners

Thank you. That is good enough for me. 

[trek27]

All arguments about costs aside it's practically impossible to asses profitability because no one knows what the market will allow DataTank to get from miners in 3-6 months from now (DTMB). In case of DTMA it's even more complicated (e.g. boards deployment timing). Nevertheless thanks for 'business case' numbers.

Though in general mining environment is extremely challenging (if not outright hostile) DataTank solution could be the way to go.

Also as potential investors we should always remember about the impact of btc valuation at the time of investment. For example, today probably everyone should seriously consider denominating into fiat all investments (shares, units, mining hardware, contracts...) paid with btc.

[webbrowser]

I have not followed closely.

Antirack, what is your relationship with DataTankMining and Allied Control?

I've also asked a few times without getting a response. Different portions of the prospectus, forum conversations suggest that  Allied Control and DataTankMining are the same entity, separate entities, supplier-customer, share the same management team, etc. What actually are the relationships?

[Collider]

DataTank is a spinoff of Allied Control.

As to who exactly Antirack is, he can tell you himself

[antirack]

Antirack, what is your relationship with DataTankMining and Allied Control?

I've also asked a few times without getting a response.

I did not see your question, apologies.

Different portions of the prospectus, forum conversations suggest that  Allied Control and DataTankMining are the same entity, separate entities, supplier-customer, share the same management team, etc. What actually are the relationships?

The post number one in this thread says the following:

DataTank Mining Limited is a technology company founded by the Bitcoin cluster engineers at Allied Control.

In post number two I personally said:

I am working for DataTank Mining.

The prospectus says:

Before DataTank Mining, the founders have invested years in building some of the largest and most efficient Bitcoin mining clusters at Allied Control, and have an exclusive agreement with the company to build and purchase DataTank mining systems at internal costs.

It is also explained in the FAQ on the web site, I spare you the copy and paste:

http://www.datatank-mining.com/faqs/partnership-with-allied-control/

DataTank Mining and Allied Control are different companies (legally). I work as VP of Engineering at Allied Control and I am a co-founder of DataTank Mining. I can use Allied Control's resources (SMD line etc) and have access to internal prices, since I have built these systems with my team. You'll also find my name (Alex) in the DTM prospectus, on the site (DTM + Allied Control), and in the company brochure (Allied Control). I have been working with Friedcat and his team since early 2013. Prior to that with FPGA developers. I have also published several articles about immersion and Bitcoin, and you can look me up on the company blog or LinkedIn.

You will find me in both of these pictures:
http://stmedia.startribune.com/images/ows_139701009277917.jpg
(where we presented the Ice-X with the Intel, SGI and 3M team)

https://twitter.com/SpondooliesTech/status/481736473197023232/photo/1
(with Asicminer (not in picture), Rockminer, Bitmaintech, Spondoolies Tech, Genesis-Mining and a few others)

I truly believe that immersing Bitcoin ASICs is a game changer. There will be chips and boards that are otherwise not possible using traditional cooling. it also changes the power profile, which means apart from having no fans that consume power, the ASIC chips, power supplies and DC circuits do a better and more efficient job, consuming less electricity. And some more.

Let me know if you have any other questions.

[webbrowser]

Thank you.