[BitFunder] IceDrill.ASIC IPO (235 Thash Mining Operation powered by HashFast)

[iCEBREAKER]

Will,

I think once you have the farm up and running, you should use a part of that 25% to take a stab at developing your own chip. There is a huge amount of incentive for this, and it will be a small chunk out the 25% when you have 300-500TH mining.

The first benefit is it will cut your costs to expand the mine, meaning for the same price you can add even more hashing power.
Secondly, it can provide a huge revenue stream, probably even bigger than mining. You can sell bulk chips, and sell your own miners.

Some off the top of the head numbers.

20,000 units X $3,000 = $60,000,000 in revenue, which 90% of it will be pure profit given the insane mark up of mining hardware.

I think it would be quite foolish to not take a stab at the manufacturing side, given the amount of money you will have from the 25% to expand the business.

Just my two cents.

Yes this, then after the ice.drill chip is done we start mining asteroids and colonizing Mars!   

[CanOpener]

That sounds like a very reasonable assessment. In your opinion, which company has the best tech to win the war?

It wouldn't make much sense to invest in their own 28nm chips, the NRE for quick tape-out/fast chip produce is around 5M.

Eventually because of the insane competition in this marketplace come January, you'll start to see the price of mining rigs hit very low prices <$5 and then the next batch will be <$2 and by the time ice.drill would have had their own chips, the price per GH will be so low ESPECIALLY for bulk wholesale buyers that the cost savings in developing their own chip wouldn't be worth the time/effort/cost.

[Visesrion]

HF provides you the chips, but the pcb design that you are gonna assenbly is also provided by them or is it your design ?

Can you also tell us the price per HF chip you got in the agreement with them ?

[Ytterbium]

Yes, it is like rocket science.  It doesnt matter how much money you have.  It is hard to compete with other engineers who live and breath bitcoin.  You cant throw any chip engineer at a problem and expect them to beat the best of engineers who have been working on this problem for over a year.  Look at cointerra's staff or HashFast's staff.  Their CTOs and/or CEOs are founders and engineers with a passion for bitcoin.  They gave up their former jobs to build bitcoin hardware.  The founder of IceDrill is not a hardware engineer (no offense to them.)

Heh, it's actually more complex then rocket science.

[Vycid]

Yes, it is like rocket science.  It doesnt matter how much money you have.  It is hard to compete with other engineers who live and breath bitcoin.  You cant throw any chip engineer at a problem and expect them to beat the best of engineers who have been working on this problem for over a year.  Look at cointerra's staff or HashFast's staff.  Their CTOs and/or CEOs are founders and engineers with a passion for bitcoin.  They gave up their former jobs to build bitcoin hardware.  The founder of IceDrill is not a hardware engineer (no offense to them.)

Heh, it's actually more complex then rocket science.

If only you knew...

The EUVL (extreme ultraviolet lithography) effort for the 10nm process node is a bigger engineering undertaking than the moon landing. And that is just one part of the process node. (Of course, there's 200 billion dollars in yearly revenue for the electronics industry and $0 in revenue for going to the moon)

http://en.wikipedia.org/wiki/Extreme_ultraviolet_lithography

[hd060053]

why did the price go to 0.0005?

Is the price of 0,0014-0,0016 too high ?

[Ukyo]

why did the price go to 0.0005?

Is the price of 0,0014-0,0016 too high ?

Ah, I think that 24hr low was due to an options trade.

[Deprived]

why did the price go to 0.0005?

Is the price of 0,0014-0,0016 too high ?

Ah, I think that 24hr low was due to an options trade.

I think (may be wrong here) that Bitfunder (same on BTC-TC) reports the strike price on option trades not Strike+Premium.  So when an option is sold with a high premium it ends up leaving a very low price in the history if exercised.

[JohnyBigs]

why did the price go to 0.0005?

Is the price of 0,0014-0,0016 too high ?

Ah, I think that 24hr low was due to an options trade.

I think (may be wrong here) that Bitfunder (same on BTC-TC) reports the strike price on option trades not Strike+Premium.  So when an option is sold with a high premium it ends up leaving a very low price in the history if exercised.

That or some idiot enjoys selling at a 75% loss. And showing options on charts is just retarded, no where in the world for any financial instrument are option trades shown in the main chart.

[Mabsark]

why did the price go to 0.0005?

Is the price of 0,0014-0,0016 too high ?

Ah, I think that 24hr low was due to an options trade.

I think (may be wrong here) that Bitfunder (same on BTC-TC) reports the strike price on option trades not Strike+Premium.  So when an option is sold with a high premium it ends up leaving a very low price in the history if exercised.

That or some idiot enjoys selling at a 75% loss. And showing options on charts is just retarded, no where in the world for any financial intrument are optioms shown for the main chart.

In that case it would have just filled the highest bid. I see people make this claim quite often. Do you guys seriously not understand how these exchanges work? That would be understandable for someone who has never used one before, but I'd expect that does not apply to people posting in this sub.

[ThickAsThieves]

What are the contingency plans if difficulty shoots through the roof?

In a sense, the shareholders are the contingency plan, no?

[will]

What are the contingency plans if difficulty shoots through the roof?

Difficulty increases seem to have increased to around ~30% and is obviously a major concern. Although beyond our control, there are some steps we could take to mitigate difficulty risk:

1. Move the infrastructure deployment closer to where the final manufacturing steps will take place, cutting logistics requirements down to the bone.
2. Leverage our buying contract(s), along with the 25% reinvestment to buy more for less on subsequent expansions. $/Gh of hardware (that is not immediately available) is currently eroding exponentially.
3. Explore other, non-degrading revenue streams which flow naturally from our current contractual base.
4. Plan for longevity. Although the cost of running these machines should not be an issue short-term, it does affect long term ROI. Part of the reinvestment money will go directly towards improving our bottom line for the long-term.
5. Overclocking based on cooling is still a hot-topic of debate. One I normally steer well clear of as the exact benefits are really all speculation at this early point. Whether or not we'll get more hashing power than we bought at relatively little extra capital expenditure (cooling infrastructure) is unknown at this stage. The research is very interesting though.

A mitigating factor beyond our control is that generation 1 ASICs (110nm+ tech) will start hitting cost/revenue equilibrium sometime around the end of the year. That's assuming spot remains at around ~$100-120. While it won't result in a difficulty decrease (I've only ever seen one of those) as the machines are shut down, we're hoping to see at least a short term slow-down in the difficulty rise as the old(er) production lines are discontinued. "hoping" is a bad word in that sentence, I used it because I can't use "planning"

The points mentioned above are not exhaustive, we're happy to take any serious suggestions on the topic on board.

[will]

A lot has been happening in the bitcoin mining world.  Cryptx decided to buy GH using bitfury to start minig in Sept.  TerraHash is going to provide hosted hashing in Nov at $6 a GH.  How will IceDrill respond?  Can you guys give us an update?  Have you guys decided what to with the extra money you have to counter the advancement of competitors?

We're keeping an eye on all the competing offers out there, but I don't feel that we should initiate knee-jerk reactions to phrases like "decided to" and "going to". I'm not saying our competitors won't deliver on their plans, they probably will I wish them all the best (for what it's worth).

We have to consider technology level, price and compatibility with our deployment plans. Drastically changing course/focus at our scale and <8 week deployment/readiness schedule would create brand new risk profiles for us, none of which are better than the one we're currently working to nullify/mitigate.

[will]

What are the contingency plans if difficulty shoots through the roof?

In a sense, the shareholders are the contingency plan, no?

I don't understand the question?

[chriswilmer]

I think we can expect a difficulty of 175-225 in November, 110 is still too low of an estimate most likely.

Good call.

[dropt]

When will the 7,000,000 million shares at 0.16 be removed if no one has purchased them?  These shares shouldn't get any dividends.

Good question.  Also, I haven't been following this closely at all.  WTF is up with the decreased farm capacity.  Title states 500TH.  The IPO shares were released on the premise of 500TH, now it's been nearly cut in half.  What's going on with this?

[Deprived]

There's some seriously strange maths going on here.

From the contract:

"The first batch is a initial sale of 14,500,000  (14.5 million) shares, which will be sold on BitFunder for 0.0014 BTC per share. "

"The second batch of 3,500,000 (3.5 million) shares and will be sold at 0.0015 BTC per share."

That's 18 million shares in total - and adds up to 25550 BTC.  Call it 25400 BTC to allow for transaction fees.

Yet the table quoted in the post above claims that with 18.5 million shares sold (500k more) only 20k BTC has been raised.  Somewhere over 5K BTC ($0.5 million) has vanished.  Have some private unpaid shares somehow accidentally got into the public section?

There's further problems with the math in the table.  Look at the 3 different examples.  They're 20k BTC, 27.5K BTC and 35K BTC - so same gap of 7.5k BTC between each set.

Now look at the hashing-power.

When capital increases from 20K to 27.5K hashing power increases from 235.3 Th to 303.03 Th - an increase of just under 68 Th.
When capital increases from 27.5k to 35K hashing power increases from 303.03 to 363 Th -- an increase of just under 60 Th.

Somehow extra hashes get MORE expensive per hash the more you buy?  That's the opposite way round to how pricing usually works on bulk purchases.

I think there's just been a cockup on BTC raised in each line (and not anything dodgy going on) - as it appears from the table that 4 million shares at .0016 raise the first extra 7.5k public BTC then 3 million more public shares (at the same price) raise another 7.5k BTC.  When in fact neither of those sizes of public share sales raises anywhere near 7.5k.  It looks like the table was changed at some point and one of the columns (BTC raised) didn't get updated when the others did.

EDIT:  Was reviewing the BTC-TC listing as it claimed to be updated.  It also has silly things like "These assets have to be sold by the 25th of October (00:00:00 UTC, 4 November 2013) before the offer expires."  In the Financial Planning section.

I appreciate you're only asking for a few million dollars - so can't really justify someone proof-reading a few paragraphs of text - but noone even skim-read it for obvious howlers like that? (which also totally contradicts the contract which claims the IPO ends in September anyway).

[dropt]

[...]

Thanks for dragging that up. 

[will]

Yet the table quoted in the post above claims that with 18.5 million shares sold (500k more) only 20k BTC has been raised.  Somewhere over 5K BTC ($0.5 million) has vanished.  Have some private unpaid shares somehow accidentally got into the public section?

We've collected 24400 BTC from public share sales in total. The 1150 BTC difference between the 25550 and 24400 represents the discounts for the block sales we did (sold at less than 0.0014). Those shares have been transferred to their respective owners on bitfunder. 4400BTC is in our possession (hasn't vanished) and will be used for deployment.

Somehow extra hashes get MORE expensive per hash the more you buy?  That's the opposite way round to how pricing usually works on bulk purchases.

The IPO share price point for reference is 0.0016 (represented by the public share payback clause). The 0.0014 and 0.0015 BTC tranches represent the early and immediate requirements for payments to Hashfast as well as setup costs. A discount was applied to the 1st two tranches to acknowledge the risk taken by those early (pre-tapeout news) investors.

I think there's just been a cockup on BTC raised in each line (and not anything dodgy going on) - as it appears from the table that 4 million shares at .0016 raise the first extra 7.5k public BTC then 3 million more public shares (at the same price) raise another 7.5k BTC.  When in fact neither of those sizes of public share sales raises anywhere near 7.5k.  It looks like the table was changed at some point and one of the columns (BTC raised) didn't get updated when the others did.

The examples in the table are for illustrative purposes and the numbers are correct. The public/private share ratio scales down/up such that a public share represents 10MH/s at the outset of the farm. The "BTC raised" in the table represents the capital expenditure for hashing power only and not the setup/logistics costs involved in deployment of the ~6 metric tons of actual mining equipment, which is covered by the 4400 BTC mentioned above.
See: https://i.imgur.com/Usl84Ah.png (as used in the BTC-TC listing), it includes the public/private share split.

EDIT:  Was reviewing the BTC-TC listing as it claimed to be updated.  It also has silly things like "These assets have to be sold by the 25th of October (00:00:00 UTC, 4 November 2013) before the offer expires."  In the Financial Planning section.

I appreciate you're only asking for a few million dollars - so can't really justify someone proof-reading a few paragraphs of text - but noone even skim-read it for obvious howlers like that? (which also totally contradicts the contract which claims the IPO ends in September anyway).

Definitely an error, thank you for pointing it out. DeadTerra has changed the "25th of October (00:00:00 UTC, 4 November 2013)" to 30th of September (00:00:00 UTC, 30 September 2013). This was my mistake, the BTC-TC listing required a reformatting/re-organisation of the IPO wording, I somehow managed to screw up that important date quite badly in the process. My most sincere apologies for this, it's a big fat glaring mistake and should have been picked up and fixed in the proof reads.

Although the BTC-TC submission was resubmitted for approval after we made the requested updates, the votes on it remain unchanged. Other than the mistake above we've had no further feedback from the voters.  The listing, for reference, is here: https://btct.co/security/ICEDRILL.ASIC (see details tab) Apologies to everyone who asked for the listing to be made there, it doesn't look like it will happen anytime soon (or at all). The forces at work there are well beyond our control.

[Deprived]

Somehow extra hashes get MORE expensive per hash the more you buy?  That's the opposite way round to how pricing usually works on bulk purchases.

The IPO share price point for reference is 0.0016 (represented by the public share payback clause). The 0.0014 and 0.0015 BTC tranches represent the early and immediate requirements for payments to Hashfast as well as setup costs. A discount was applied to the 1st two tranches to acknowledge the risk taken by those early (pre-tapeout news) investors.

Think you misunderstood my point here.

I wasn't talking about later shares being more expensive - I understand why you did that (though I don't necessarily agree with the reasoning).

I was talking about you getting less hash/USD from Hashfast.

You get more hashes (over 10% more) per BTC (and in absolute terms) according to the table for the next 7500 BTC you raise than for the final 7500 BTC after that.