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

[LeanSixSigma]

Well duh!  Oddly enough, this data is collected in respect to capacitors exposed to air, not Novec.  To be more explicit, it doesn't suggest that soaking capacitors in Novec or chilled donkey piss will extend MTBF.  With no published data re. electrolytic cap MTBF in Novec 7000-series fluids is available, it's nonsensical to assume that a cap will live as long in cool Novec as it will in cool air, the environment for which it was designed.  For someone so fond of studies, you certainly have a lot to learn about controls

The studies are very clear: Higher temperatures -> capacitors will fail faster. The Army Research Laboratory confirms 2-phase cooling does not lead to any hot spots like with air cooling: "Along with the other previously mentioned potential benefits of two-phase cooling, this can provide a uniform heat flux and temperature across the electronics", while the previous other quote mentioned heat shields etc. So argue against the high temperature correlation to failure rate at first, because that is pretty much a well-studied fact.

The US military has performed many long-term studies on 2-phase immersion cooling, but not all are public. More evidence below. One public study I know is a 3M engineer who has completely immersed a CRT TV in fluid and it was still working after 10 years. Not much of statistical significance. But the two mining clusters of Allied Control were running without ANY SINGLE of your mentioned capacitor issues. Literally in the hundred thousands of capacitors running non-stop (equally high number of observations).

Without any study on your side, you can simply not claim that MTBF is shorter. Your claim has as much merit as that if you buy a lottery ticket that you will definitely win the big jackpot. It may be within the theoretical range of possibilities, but since you simply make a wild claim and can't prove it, it's therefore again, a pure false conjecture.

Does your DataTank system have the ability to filter out electrolytes from Novec?  Feel free to expand, but try to keep it terse...

You asked for it: Yes.

The maker of extraordinary claims is the one who needs to provide collaborating evidence.  You have provided none.  Personal accounts, "I talked to a d00d who said an alien anal probed him," doesn't constitute such evidence.  Lrn to science pl0x.

That's why I have provided a data center study which wasn't even limited to China, but showed creep corrosion even with banks, ISPs, etc. You failed to notice it for the third time or so. But I'm refraining like 2112 to call people illiterate - I just don't think that anyone deserves that. So it's maybe deliberately omitting/ignoring the earlier referenced study just to desperately try to win the argument? The personal accounts are only backup thrown in for free like candy at the reception desk.

But what you have not provided yet are any sources to dispel my claims. You just simply make a false and unfounded claim. Again, pure false conjecture.

Electronics specifically designed for such cooling.  The US army has also purchased multi-thousand dollar toilet seats, and experimented with using bats as vectors for dispersing incendiaries over Japan...

Yes, and not every experiment is useful. But 2-phase cooling has been successfully implemented in UAV drones. Global Hawks are in service since 2005 and Predators since "early 1990s" already. Sounds a bit more than just an experiment to me. This shows as well that drones in service using 2-phase cooling have a much longer service record than you might have expected. Try arguing against that with your capacitor point.

Please try to be concise.  The shotgun approach you seem to favor is getting tiresome.
ty

You see, it's sometimes necessary to make a claim wholesome, back it up with reliable sources, etc. to make it strong. I'm not sure whether you read my posts completely, but it appears that you at least don't read the referenced sources carefully - otherwise you would not come again with creep corrosion data center and still not being able to provide any proof. You really like avoiding providing anything to substantiate your arguments, because you simply do not seem to have that - exactly that's the reason why I can knock out your unfounded/wrong claims so easily. This time was shorter, right? Thanks to you I had the opportunity to build my argumentation base already and can now more conveniently refer back to already written stuff. You are proud contributor to a 2-phase cooling FAQ dispelling myths, promoting the technology more with every single post.

[NotLambchop]

...

Does your DataTank system have the ability to filter out electrolytes from Novec?  Feel free to expand, but try to keep it terse...

You asked for it: Yes.
...

ORLY?  Please elaborate on the process used, and the types of electrolytes it is capable of removing/neutralizing.  Here, feel free to be as expansive as you wish
ty

*Though do try to keep your reply focused.

[LeanSixSigma]

ORLY?  Please elaborate on the process used, and the types of electrolytes it is capable of removing/neutralizing.  Here, feel free to be as expansive as you wish
ty

*Though do try to keep your reply focused.

I would really like to do that to show off, especially because we have worked very hard on our cleaning system and are kind of proud - but we are preparing to protect it. So until then, we can't reveal any details on that to the public yet (basic procedure). If you like, then I can provide you the contact to our law firm with the relevant US attorney to vouch for that personally.

You could as well make an offer and buy it. It's up to you then to publish it here if you are so keen on it. It won't be cheap though. Ask for any mining hardware manufacturer to give you their ASIC mask because you don't believe their performance claims - similar story. Building a small scale experimental 2-phase immersion cooling system is one thing. Building it commercially feasible for heavy duty industrial use is another - that's where the key expertise sits and why 3M is so keen on working with Allied Control, since systems they have worked on were more experimental or one-off only, without continuous operation for long time in large scale - the military has such applications, too.

So I assume that the non-reply to the US Army Research Laboratory study and the successful use in UAV drones, clearly stating the cooling superiority and mission-critical long-term reliability of 2-phase immersion cooling over air and water cooling in high density electronic applications (you claimed the contrary before), can be assumed as you having no counter-argument with sources any more, and therefore admitting to it as key benefit? Let me return the favor: Please keep your reply focused - A clear Yes or No, if No then prove it with sources. No distraction on this one.

[NotLambchop]

ORLY?  Please elaborate on the process used, and the types of electrolytes it is capable of removing/neutralizing.  Here, feel free to be as expansive as you wish
ty

*Though do try to keep your reply focused.

I would really like to do that to show off, especially because we have worked very hard on our cleaning system, but we are preparing to protect it. So until then, we can't reveal any details on that to the public yet (basic procedure). If you like, then I can provide you the contact to our law firm with the relevant US attorney to vouch for that personally.

You could as well make an offer and buy it. It's up to you then to publish it here if you are so keen on it. It won't be cheap though. Ask for any mining hardware manufacturer to give you their ASIC mask because you don't believe their performance claims - similar story. Building a small scale experimental 2-phase immersion cooling system is one thing. Building it commercially feasible is another - that's where the key expertise sits and why 3M is so keen working with Allied Control, since all systems they have worked on were more experimental or one-off only without continuous operation for long time in large scale.

So I assume that the non-reply to the US Army Research Laboratory study and the successful use in UAV drones, clearly stating the cooling superiority and mission-critical long-term reliability of 2-phase immersion cooling over air and water cooling in high density electronic applications (you claimed the contrary before), can be assumed as you having no counter-argument with sources any more, and therefore admitting to it as key benefit? Please keep focused - A clear Yes or No, if No then prove it with sources. No distraction on this one.

Lol, "we have it, but can't tell you because secrit"  Try harder, this one's been pretty much played out here on Bitcointalk.

Re. US army:  phase change cooling works exceptionally well, it worked for thousands of years, it takes place every time you splash some water on a hot rock.  It worked surprisingly well in hit and miss engines of the early 1900s, which simply allowed the coolant (water) to boil away.  The physics are not in question here.



OTOH, extrapolating from this that common electrolytic capacitors would function reliably is pure nonsense.

The question is what is being cooled, and with what fluid.  US Army used purpose-designed components in vehicles which were to be torn down, tested and serviced at short, regular intervals--approximately the opposite of what is required of a Bitcoin mining farm.

[Jutarul]

Would somebody please make an executive summary of the current trolling (200-300 words)? Thanks.

[NotLambchop]

Sure.  LeanSixSigma is trying to sell the rubes of this forum shares of a pricey 40-footer full of unicorn tears.
He claims that it will solve the fundamental problem faced by Bitcoin miners:  "creeping corrosion."  You do not know about this creeping corrosion because the miners are too proud to talk about it, but LeanSixSigma knows it's there, and it's super srs.

[OgNasty]

Sure.  LeanSixSigma is trying to sell the rubes of this forum shares of a pricey 40-footer full of unicorn tears.
He claims that it will solve the fundamental problem faced by Bitcoin miners:  "creeping corrosion."  You do not know about this creeping corrosion because the miners are too proud to talk about it, but LeanSixSigma knows it's there, and it's super srs.

I don't think it solves a problem that is currently plaguing miners.  However, with ASIC development inevitably slowing as technological limitations begin being reached, corrosion could become an issue as mining equipment stays profitable for longer.  Exploring this angle couldn't be a bad thing, although I don't think we're there yet.  Thinking that development will continue to make miners useless every year forever would not be smart.

[LeanSixSigma]

Re. US army:  phase change cooling works exceptionally well, it worked for thousands of years, it takes place every time you splash some water on a hot rock.  It worked surprisingly well in hit and miss engines of the early 1900s, which simply allowed the coolant (water) to boil away.  The physics are not in question here.

Thanks - so that's settled once and for all. NotLambchop acknowledges that 2-phase immersion cooling works exceptionally well.

OTOH, extrapolating from this that common electrolytic capacitors would function reliably is pure nonsense.

The question is what is being cooled, and with what fluid.  US Army used purpose-designed components in vehicles which were to be torn down, tested and serviced at short, regular intervals--approximately the opposite of what is required of a Bitcoin mining farm.

Check with Northrop Grumman - there is a white paper confirming that 2-phase cooling promotes the use of "commercial grade electronics", specifically referring to tantalum and ceramic capacitors, where otherwise more durable components have to be used. And regardless of that - once again, you are conveniently omitting my other claim with hundreds of thousands of completely normal capacitors in long-term use in two large scale industrial, 2-phase immersion cooled Bitcoin mines. Not a single capacitor failed. Not an extrapolation, but actual real-world proof. This combined is still stronger than your wild and completely unfounded claim pulled out of thin air that capacitors have a lower MTBF under 2-phase cooling.

He claims that it will solve the fundamental problem faced by Bitcoin miners:  "creeping corrosion."  You do not know about this creeping corrosion because the miners are too proud to talk about it, but LeanSixSigma knows it's there, and it's super srs.

Not true. I have never claimed that it was the fundamental problem faced by Bitcoin miners in general. To me that's only a potential side effect when looking at the open air facilities - the key argument on more cost-efficient cooling has been already won above.

I have even posted that there are many other Bitcoin miners not facing that problem (proper building, better air, etc.), but that there is a risk of that occurring for open air facilities, next to other pollution, dust and humidity (which was again omitted), having potential effect on reliability. I have never ever claimed that I know for sure it's there and posted that for clarification earlier, but only that there could be the possibility/risk, supported by the supplied reliable studies on similar or even better scenarios (data centers and other buildings). So a logical reasoning would be that worse scenarios (complete open sides, exposed to nature, high air flow supplying with more contaminants), might have a bigger effect on reliability, and only what the consequences could be "IF" it happens. Again, falsely claiming things which I never wrote. Looks to me like grasping for straws because you're running out of arguments.

Would somebody please make an executive summary of the current trolling (200-300 words)? Thanks.

NotLambchop and others made various unfounded claims, trying to question the sense of 2-phase immersion cooling. But essentially all claims made so far are either apparent irrelevant diversions or just simply outright wrong claims. No reliable source has been provided so far from their side to support their arguments. I have on the other hand provided multiple scientific studies to back up my claims and those could still not be dispelled. Btw.: It's interesting that NotLambchop didn't care to deny when I "seriously" asked whether he gets paid for these posts - I would have jumped on it and denied clearly if it wasn't the case. If that was true then it might shed another light on the motivation/background of trolling here.

But what counts most to me is that NotLambchop eventually doesn't deny the findings of the US Army Research Laboratory on overall more reliable and cost-efficient cooling than air and water cooling for high powered electronics, and agreed that 2-phase cooling "works exceptionally well". It's the key and main argumentation line. Everything else are just less relevant branches/forks of side-arguments.

[IPO Magic]

...
Check with Northrop Grumman - there is a white paper confirming that 2-phase cooling promotes the use of "commercial grade electronics", specifically referring to tantalum and ceramic capacitors, where otherwise more durable components have to be used.
...

Lol, so what's your area of expertise, if you don't know that tantalum and ceramic caps are not electrolytic capacitors being discussed here?  Tantalum caps are much more expensive per mF, and ceramics are only good for pico range.  Learn to electronics.  
BTW, stop equating phase change cooling with Novec.  The mill engines I've mentioned are boiling plain old water.
Flaunting your ignorance may feel liberating, but it's absolutely painful to watch

[Jutarul]

Sure.  LeanSixSigma is trying to sell the rubes of this forum shares of a pricey 40-footer full of unicorn tears.
He claims that it will solve the fundamental problem faced by Bitcoin miners:  "creeping corrosion."  You do not know about this creeping corrosion because the miners are too proud to talk about it, but LeanSixSigma knows it's there, and it's super srs.

I don't think it solves a problem that is currently plaguing miners.  However, with ASIC development inevitably slowing as technological limitations begin being reached, corrosion could become an issue as mining equipment stays profitable for longer.  Exploring this angle couldn't be a bad thing, although I don't think we're there yet.  Thinking that development will continue to make miners useless every year forever would not be smart.

Correct. Solving longevity (>1-2 years) problems of equipment in the current environment is a bad allocation of time and investment capital. Competitive edge is realized in reliance (which includes sourcing electronic components) and deployment speed. And based on what DataTank mining is proposing that is a strong point about them.

The discussion about phase conversion based cooling is really a discussion about how to get the heat transported into the environment effectively. The current strategy to put equipment into a wind tunnel only scales so far... there is a point at which you get diminishing returns (av. air temp goes up, inc. air speed reduces conductance) - it would be interesting to know what the theoretical limits of air cooling are. With those limits one can get an estimate of the minimum real estate cost per kW hashing power.

[LeanSixSigma]

Lol, so what's your area of expertise, if you don't know that tantalum and ceramic caps are not electrolytic capacitors being discussed here?  Tantalum caps are much more expensive per mF, and ceramics are only good for pico range.  Learn to electronics.  

That's exactly why I have posted: "regardless of that - once again, you are conveniently omitting my other claim with hundreds of thousands of completely normal capacitors in long-term use in two large scale industrial, 2-phase immersion cooled Bitcoin mines. Not a single capacitor failed. Not an extrapolation, but actual real-world proof." - it is the relevant argument here which you can still not counter.

I've posted my area of expertise already, but we know nothing from you except your experience as five year old fiddling with resistors and "playing with scraps of wire I stripped with my little teeth".

BTW, stop equating phase change cooling with Novec.  The mill engines I've mentioned are boiling plain old water.

I was referring to the very same US Army Research Laboratory study and it actually analyzes single phase water cooled systems (water inlet temperatures): "Most silicon power electronics have a full-power operating temperature limit of about 125 °C [...] These high inlet temperatures limit the electronic device temperature increase to only 25–45 °C above inlet temperature, reducing electronic cooling potential and requiring exceptional cooling performance to dissipate high-heat-fluxes and prevent overheating." and goes on to compare with a superior 2-phase cooling system which "was implemented using HFE-7100 dielectric coolant fluid to safely dissipate 150-200W/cm2" - very comparable or even higher than most low nm crypto mining ASICs, incl. commercial GPUs for Scrypt. And that's by far not the limit yet.

So yes, they were using in that particular test single phase water cooled systems vs. Novec 7100 and found 2-phase cooling to be superior with significantly less pumped water in secondary stage (less cooling electricity = less costs) without letting the chips overheat. No conjecture, referenced proof.

It doesn't matter whether you claim that you only meant water 2-phase cooling is more effective and not Novec. It's the same physical principle generically for ALL 2-phase cooling systems and have already quoted that earlier out of the same report:

"First, a fluid’s latent heat of vaporization has a fundamental limit that can be two orders of magnitude larger than the specific sensible heat of single-phase liquid cooling. Thus, the boiling effect provides the possibility of increased heat absorption per unit volume of fluid and higher heat acquisition effectiveness (i.e., the amount of heat absorbed by a unit of flow relative to its maximum theoretical capacity)."

The mill engines I've mentioned are boiling plain old water.

So let me get this straight: IPO Magic never posted anything about "mill engines" (= military engines), but NotLambchop did instead. I've checked every single post of IPO Magic in this thread to confirm. I've already thought about that possibility, but now we've got clear evidence: That means that the person behind IPO Magic and NotLambchop is one and the same and is deliberately using multiple handles to create the false illusion of multiple posters bringing up arguments against 2-phase cooling, one of which not denying to get paid for posting. That's just very low and unethical if one needs to resort to such tactics and can still not make any significant progress. At least to me, any kind of credibility of these two posters is pretty much confirmed to be down the drain with this (likely stating the obvious for other more credible posters who had the honor to be insulted by IPO Magic and NotLambchop).

[NotLambchop]

...I've posted my area of expertise already...

What is it?

I was referring to the very same US Army Research Laboratory study...

Is this a published study?  Title? ISBN? Link?

So let me get this straight: IPO Magic never posted anything about "mill engines" (= military engines),

Lol, Google "mill engine" to  savor the full extent of your ignorance

but NotLambchop did instead. I've checked every single post of IPO Magic in this thread to confirm. I've already thought about that possibility, but now we've got clear evidence: That means that the person behind IPO Magic and NotLambchop is one and the same and is deliberately using multiple handles to create the false illusion of multiple posters bringing up arguments against 2-phase cooling, one of which not denying to get paid for posting. That's just very low and unethical if one needs to resort to such tactics and can still not make any significant progress. At least to me, any kind of credibility of these two posters is pretty much confirmed to be down the drain with this (likely stating the obvious for other more credible posters who had the honor to be insulted by IPO Magic and NotLambchop).

Wait, you're saying NotLambchop is a sock?  But if Lambchop is a sock, how could NotLambchop also be a...  Ohhhh....  He probably thought he was so cunning with that "Not", but there's no fooling a clever feller like you!  No sir!



*I've been accused of being Mircea Popescu (Romanian exchange owner), Eduardo de Castro (HashFast), crumbs (of Active Mining fame), MikeMikeMike and Puppet (PETA), and countless other bogeymen.
If it helps you to think me any or all of the above, or that I'm being paid for my posts, I won't ruin it for you.  Makes you that much more amusing to deal with.
ty

[Bicknellski]

Sixsigma - 3

Lambchop - 0

What is the score now?

[Collider]

Yeah, he comes up with some arbitrary problems he seems to deem to hard to overcome and claims the guys haven´t thought the system through.

Obviously nobody is going to hand him the exact plans to build his own, so he is spamming the forum and flinging shit like an ape.

I hope he remembers his words when in 3-6 months everything is set up and working fine.

[NotLambchop]

Having everything "set up and working fine" tells me nothing about improving MTBF, commercial viability (err... unless you consider selling unicorn dreams to ignorant bitcoiners "proof of viability"), or usefulness of this Goldbergian contraption (Rube Goldberg's machines did exactly what they were designed to do).
That said, I'm guessing nothing (other than a few burnt "investors") would come of this.

As far as me wanting to scope out DataTank's secret sauce?  What DataTank is doing has been done multiple times before, there's nothing mysterious about the basic process.  It hasn't been shelved because of insurmountable technical problems, but simply because cost/benefit makes no sense.

As far as complexity?  This is a functional phase-change cooling setup.  Not rocket surgery.  Shell out $350 for a gallon of unicorn tears, and have a blast!

 
https://bitcointalk.org/index.php?topic=7216.msg7644407#msg7644407
(it is wikid fun)

[LeanSixSigma]

Having everything "set up and working fine" tells me nothing about improving MTBF, commercial viability (err... unless you consider selling unicorn dreams to ignorant bitcoiners "proof of viability"), or usefulness of this Goldbergian contraption (Rube Goldberg's machines did exactly what they were designed to do).

As posted before already, we have built multiple profitable and large scale 2-phase immersion cooled Bitcoin mines and received various media coverage to prove it.

As far as me wanting to scope out DataTank's secret sauce?  What DataTank is doing has been done multiple times before, there's nothing mysterious about the basic process.  It hasn't been shelved because of insurmountable technical problems, but simply because cost/benefit makes no sense.

As far as complexity?  This is a functional phase-change cooling setup.  Not rocket surgery.  Shell out $350 for a gallon of unicorn tears, and have a blast!

Look for my earlier post on why it didn't take off before: The fluid price as ratio over low density hardware was not so good in the past. But especially cryptocurrency mining hardware has changed that and it's still continuously increasing. Further, as pointed out earlier, the price for e.g. tens of thousands of pounds is different - another key advantage as 3M Technology Partner vs. others. As mentioned earlier, we have set up profitable mines = cost/benefit made sense.

Yes, the basic principle is simple and that's why we like it so much. But again, google for Dunning-Kruger effect: Go ahead and build one yourself without losing too much fluid. Or apply for a position as professional football player - the basic principle/rules are simple and it looks easy on TV, right?

Experimental and commercially viable systems are two different things. We have invested a lot of money into R&D and many tens of thousands of hours over years to get this far and have therefore an immense head start (like fluid loss prevention). 3M chose to work with us as official 3M Technology Partner for 2-phase immersion cooling technology, since we were able to build multiple commercially viable large scale commercial systems.

Having said that, we think it's good that there are more enthusiasts jumping on it and congratulated them already. smracer recognized the superior 2-phase cooling abilities, confirming what I've mentioned earlier. So at the same time you are actually making a point that there are more people believing in this technology. Thanks! And once again I appreciate that you give me the opportunity to provide any interested reader here with more info on why we think that we have a competitive advantage.

***

I try to keep this short, since it's OT - but I think it's beneficial to this thread enough to post it to address the troll problem here:

Guess I searched a bit around like jimmothy did, and found another alt account "Ask Ken About Love". That account together with IPO Magic and NotLambchop (and apparently a few others) troll almost exclusively threads in the securities group. You use multiple of those alt accounts within the same thread in alternation to apply the same false illusion of multiple persons having the same message. There are many other posters who got pissed off in very same fashion - so really good call on jimmothy to open a thread on that: https://bitcointalk.org/index.php?topic=677833.0

Will it help to discourage you from using those tactics? No idea, maybe not. But whenever you post something, we can point out this troll alert for newbies to this and other thread and the entire credibility of your post will be questioned, no matter what you write. It could be like some kind of DOS defense, where the effort to dispel your unfounded claims will be minimized for frequent and low effort repetition from our side.

[Bicknellski]

Any updates since the last update of 23 Jun 2014?

The opportunity here is really interesting but what are some of potential projection on the returns?

VC's would expect a better than "average" return on this sort of investment. Distinguished Innovator Lecture - Adam Grosser

I know with mining it is a crap shoot of sorts given the price of BTC etc but the margins right now are small and returns for mining are very low even on a massive scale is that not correct?

----

OT - I think this troll and his alter egos are actually very good for this security offering.

1. It shows the professionalism of the team and further provides mountains of evidence to support the claims of Data Tank Mining.
2. It shows clearly that the design implications for chips and boards is increasing density and we are already seeing the limits of heat sinks.
3. It correctly questions the viability of air, oil or water blocks moving forward and exposes the weaknesses in the other systems.
4. It attacks the weakest point in the Data Tank Mining offering that cost to cost comparisons of DTM to Massive Chicken Coop (MCC) designs.
5. It is a comparison of apples to oranges, MCC to DTM, but we want to have an itemized line for line of these and a pretty spreadsheet of it.

[novusordo]

What is your plan if Havelock disappears off the map? A fallback needs to be in place for this.

[jimmothy]

What is your plan if Havelock disappears off the map? A fallback needs to be in place for this.

+1

It would be nice to know if shareholder records will be kept outside of havelock, as well as a backup plan for a catastrophic failure of the exchange.

[Bicknellski]

Yeah, he comes up with some arbitrary problems he seems to deem to hard to overcome and claims the guys haven´t thought the system through.

Obviously nobody is going to hand him the exact plans to build his own, so he is spamming the forum and flinging shit like an ape.

I hope he remembers his words when in 3-6 months everything is set up and working fine.

He has exhausted every avenue now and DTM guys have skillful and without malice answered everything to my satisfaction. He and his posse of alts are on ignore. I am more keen on the what DTM thinks will be the returns for investors? Give we all know the margins for mining are slim even in massive farms is there realistically any potential return for basically being VC funders without the benefits of 'ownership'?