Official BITMINE CoinCraft series 28nm ASIC miners thread

[drbell]

Ohhh No ...   http://thegenesisblock.com/mining/a/d1493905be  

If you want to see what the delays have cost you then change the mining start date to Feb.

If you want to get really depressed see how the CPP is gonna help, set mining start date back to March and set Hashing Rate to 1100 (10% CPP) and press Calculate, change it to 20% CPP (1200), change it to 30% CPP (1300), only when you get to 60% CPP will the figures show a profit.

[Schinder75]

For all customers in March batch its the same. They will not be able to make any profit.

I did calculations and it shows me 2.8 BTC until the end of the year (excluded electricity) with a 600GHs CoinDesk and difficulty increase of 20% per 11 day timeframe.

Invest : 3.500$
Mining : 2.800$ (with BTC at value 1000$!)

So if bitmine will not compensate this, every customer will make a big loss...;-(

[alfabitcoin]

For all customers in March batch its the same. They will not be able to make any profit.

I did calculations and it shows me 2.8 BTC until the end of the year (excluded electricity) with a 600GHs CoinDesk and difficulty increase of 20% per 11 day timeframe.

Invest : 3.500$
Mining : 2.800$ (with BTC at value 1000$!)

So if bitmine will not compensate this, every customer will make a big loss...;-(

the difficulty wont increase 20% over this year for sure.

Miners mine btc, not fiat.

[ujka]

For all customers in March batch its the same. They will not be able to make any profit.

I did calculations and it shows me 2.8 BTC until the end of the year (excluded electricity) with a 600GHs CoinDesk and difficulty increase of 20% per 11 day timeframe.

20% per 11 days? Your calc ended with 10'000 petahash of total network hashrate in december? Now it's at 20.

[Schinder75]

For all customers in March batch its the same. They will not be able to make any profit.

I did calculations and it shows me 2.8 BTC until the end of the year (excluded electricity) with a 600GHs CoinDesk and difficulty increase of 20% per 11 day timeframe.

20% per 11 days? Your calc ended with 10'000 petahash of total network hashrate in december? Now it's at 20.

I did calculations for this year until december. The last jumps were about 20% each. I cannot see why this should stop or decrease. What makes you hope that diff will not increase with these steps? KNC datacenters around the globe a.s.o.

[masterOfDisaster]

I admit that I'm neither happy with that delay (and strangely I still hope to get my miner) nor with most part of the communication by bitmine.ch.
Nevertheless I don't want to leave the expectations about difficulty increase uncommented.
A continuous increase by a fixed percentage can't seriously be expected for an unlimited time.

This has several reasons. The conceptual reason is that there can't be unlimited growth in limited environments.

To give a more practical reason: the initial cost per hashrate and the operation cost of miners limit the growth. We are still in a phase of tremendous technological progress. But the bitmine products might be close enough to future products (in terms of initial and operating cost) that very soon the incentive for buying BTC at exchanges will be even higher than it is right now.

This will limit the hash rate growth.

Another reason that will limit the hash rate growth is the fact that you can't sustain a fixed percentage in growth by adding fixed amounts of hash rate. I know that the hash rate growth still is not linear, but it is likely to transform to linear growth if any.
When I purchased my miner at bitmine in October 2013 I asked myself whether it'd be better off to purchase BTC instead. The hope for a better return by purchasing this miner lead me to the buy. I knew about the risks. I made the wrong decision. Even if it had been delivered on time, I might have been better off buying BTC.
People buy miners and crypto currencies because of several reasons. One is to make profit.
This will have an effect on the hash rate growth.

It will become harder and harder to make profit by purchasing and operating mining devices.

Or to say it clearly and precisely:
It will become more and more obvious that making profit with hardware might be limited to the manufacturers, sellers, resellers and electric energy suppliers.

Which gives me (little in fact, but who knows...) hope, that once the bitmine devices are delivered and given the power-save mode works as announced (while the turbo mode isn't...), these miners might be able to run for a while.
The hash rate will not grow by 20% each 11.2 days forever, but nobody knows when and how much it slows down.

I don't expect BTC ROI. For that each of the 200 GH/s blades had to make more than 10 BTC.
But I'm still hoping for USD ROI. That might be possible - especially if the BTC prices rises

[ujka]

Nice post. I wish I could type that fast. And think that fast in English. Google translate isn't helping much.

Operational cost.
Just electricity requirements for a 10000 petahash network with ASICs at 1J/GHash (even KNC announced 20nm chip isn't much better) is 10GW. Take $0.15/KWh, BTC has to be over $10'000 to cover this.

[IronEagle]

Why are most of you getting nervous?? Is it because of the low BTC price?? I think if we had 1000 USD per BTC you would be much cooler.

Every company delivering miners on the market has delays. KNC, Cointerra, Bitmine etc. This is normal if you ordering special maschines. If a company would order a special construction maschine it could be delayed, if you order a car it could be delayed. I´m still waiting for my cointerras. So what? It is normal...

I think Bitmine has the best customer protection at the moment. KNCs Plan B could also be good, but I don´t know how they would build these much miners to compensate a late delivery for all Neptune customers. Of course a company could get bankrupt. Every company could get bankrupt, even Apple. So this is normal. If you getting in touch with BTC Business you need to have titanium cojones. Like if you trading future contracts on the exchanges. If you don´t have them you are wrong. If you don´t have the money or if you need a credit for this you are absolutely wrong!!! Sorry for this, but it´s true. ("Who has a lot of money, can speculate.
Who has little money, must not speculate. Who has no money, has to speculate." - Andre Kostolany)

And why don´t you order at KNC, Cointerra and Bitmine at the same time and hedging. If one is delayed or one company would get bankrupt you would have your miners from the other companys. Or just buy in-hand-miners if you don´t want to have any risk? But you have to pay much more for these then doing a pre-order.

BTC and buying miners is like every other investment or buying. You always have a risk. If you don´t want to have this go to your local bank and get a bankbook.

[masterOfDisaster]

Nice post. I wish I could type that fast. And think that fast in English. Google translate isn't helping much.

Operational cost.
Just electricity requirements for a 10000 petahash network with ASICs at 1J/GHash (even KNC announced 20nm chip isn't much better) is 10GW. Take $0.15/KWh, BTC has to be over $10'000 to cover this.

Allow me to add some numbers to the operational cost.
Currently the Bitcoin network has a total hash rate of approximately 20 PH/s (=20.000.000 GH/s).
Assuming that all mining devices did only consume 1J/GH or 1 Watt/GH/s (which is not true as thera are lots of less energy efficient devices in the field), this would result in a total electric power of 20 MW.
Assuming further that these 20 MW generate no additional costs (e.g. for cooling), they'd still sum up to 480 MWh each day.
Given a price of 0.15 USD/kWh or 150 USD/MWh this leads to operational costs of (I can safely say at least due to the optimistic assumptions)
150 USD/MWh * 480 MWh = 72.000 USD.
If my math isn't broken sustaining the current network hash rate costs roughly 72.000 USD per day
At the same time there are (assumption: stable hash rate) approximately 3600 BTC generated per day. At current exchange price of roughly 650 USD/BTC the value of the daily generate BTC is ~ 2.340.000 USD.

That's why there is still hope for getting ROI by buying and operating mining devices

[matthewh3]

Gratz on 1GH/W.

I was wondering how you did it until I realized it's 12x A1 chips per board.

210GH/12 = 17.5Gh per chip @ 1GH/W.

For the guys asking about overclock 35GH per chip has been reached by Technobit Hex8A1, but you need 4500rpm fans and huge heatsinks.

Potentially 12x 35GH x 5 = 2100GH but it would have to outside the case.

Correct math, but wrong assumptions.

What has been shown on the video is a fully populated CoinCraft Desk, which has 5 boards with 8 chips each running at 800MHz and reaching 25GHps. That's exactly the 1THps you can see.

The updated Rig board were added 2 chips to be able to provide the nominal hashrate with under-clocked chips. With that, the chips can run below 700MHz, resulting in controllable system temperatures


As for the over-clockabilities in general: with extreme cooling I pushed the chip to 1.15GHz / 36.8GHps (47%+). For that, the voltage needs to be raised from 0.85 to 1.1V, resulting in the degradation of power efficiency of 0.7J/GH to over 2J/GH - which is in line with the expectations based on physics (consumption proportional to (clock^2 * voltage)). Bottom line: while chips are over-clockable as advertised, do not expect to reach the full potential with the CoinCraft products. For one, getting 50% more hashrate requires 3x more power - which the PSU is not dimensioned for; consequently for the other you can not get 3x more heat out of a closed box unless you submerge it or live close to the poles.

For those who ordered the CoinCraft rig with extreme overclocking in mind, I frankly tell you to better expect max 20% above nominal speed. You might top that if you modify cooling and power yourself, but then Bitmine's turnkey products that plug-and-mine are maybe not the right choice for you.

So are you saying 2.8(TH/s)@3kW as advertised for the rig is not possible?

[ujka]

if you can't get riot with the BTC at the price you bought the miner, it would be much smarter to just buy BTC instead of the miner.

yea, we all know that now. but can't go back in time.

[masterOfDisaster]

@masterofdisaster
quickly add one more 0 -- $2'340'000  per day  
And with 20% increase there are more blocks per day...

Thanks - I've corrected it. Sometimes I should properly read the output of my calculator^^
And you are right about the 20% that currently lead to roughly 4500 BTC issued by coinbase transactions each day...
But the stable hash rate was just one of the assumptions. There are lots more and I strongly recommend to use this caluculation rather for getting a feeling for the dimensions than to nail me down on the exact numbers

[iglasses]

Hey guys

I am wondering if it is worth to buy coincraft desk at bitmine.ch. now when I am reading this It think it is waste of money, you are all caltulating that btc shoud rise a lot if you want to break even!?

So... is not then better to buy BTC than this asic miner?

Is any body even had good expirie ce with this company?! That he get his miner on time?

I see this company is totly not serious

You might wan to follow these topic for the coming 2 weeks and see the result?

I am following this topic from new year... And I got the filling that bitmine.ch is the biggest scam ever... Even if they sent miners now, difficoulty of bitcoin mining will be so high that they will loose a lot of money...

And if I have exact informatoin they get for a delay of 20 days 10% biger hashrate...but Difficoulti of  bitcoin is rising a bout 25-30% in 14 days... If that is not a scam I do not now what is!

I don't understand why this even matters to you.  Even if you believe everything they are saying if you didn't pre-order one of these machines months and months ago you can't get one anyway.

[BTCog]

Ohhh No ...   http://thegenesisblock.com/mining/a/d1493905be  

If you want to see what the delays have cost you then change the mining start date to Feb.

If you want to get really depressed see how the CPP is gonna help, set mining start date back to March and set Hashing Rate to 1100 (10% CPP) and press Calculate, change it to 20% CPP (1200), change it to 30% CPP (1300), only when you get to 60% CPP will the figures show a profit.

I don't think this is an "Ohh No" situation at least it shouldn't be, if you are playing this game.

BTC /USD cost predicates vs cost of electricity everything.
If BTC drops to 0 then mining will not be profitable, unless the impossible condition of getting paid to use electricity comes around.

any graph projecting computational hash rate to continue increasing exponentially is laughable.

As mentioned in another posting here - the TOTAL amount of electricity required to power TOTAL hashing, at some point, becomes cost prohibitive.
I'm no genius, but as some point before that, consumers stop buying miners and at some point before that, manufacturers stop producing 'consumer'  miners.

BTC value MUST* increase in order for mining, mining manufacturing to be viable.

*this is after 20nm or smaller manufacturing viability.

Difficulty and hashing will tap out - continuous expansion is not economically sustainable.

SMART Manufacturers know this and are:
1. sending public messages out to other manufacturers letting them know that manufacturing cannot continue indefinitely
2. building data centers for mining... this will be the only place mining will take place, perhaps at 20nm.. perhaps at 16 or 14 (16's half-node).. by which point the average joe will be locked out of mining (should btc continue to increase in value).

If you believe in BTC, grab a miner, or buy BTC.

I think 28's may have a longer life than many think - and may, in fact be more valuable.
20's even longer, and more valuable... if you can get your hands on one.

[georgem]

If you believe in BTC, grab a miner, or buy BTC.

I think 28's may have a longer life than many think - and may, in fact be more valuable.
20's even longer, and more valuable... if you can get your hands on one.

This. I want a miner because I want BTC to succeed.

Aslong as they generate more money than they consume energy,  I will let the machine run.

What do you say? How will I get back my ROI?

Depends how much I payed for the bitcoins, and when exactly I sold them to bitmine.

Some guys were lucky. :-)
Some guys should have better kept their BTC. I am sorry for them. :-(

[Puppet]

which is in line with the expectations based on physics (consumption proportional to (clock^2 * voltage)).

..that should be clock * voltage²

[zefir]

which is in line with the expectations based on physics (consumption proportional to (clock^2 * voltage)).

..that should be clock * voltage²

Thanks P. for the correction, fixed that in the post.

I did not notice since I used similar increase factors for both parameters, resulting in same results with correct and wrong formula.

[Gaglam]

sooo... it's time for the truth.


where is the very first tracking nr. ?

I think there is none

[Biomech]

For all customers in March batch its the same. They will not be able to make any profit.

I did calculations and it shows me 2.8 BTC until the end of the year (excluded electricity) with a 600GHs CoinDesk and difficulty increase of 20% per 11 day timeframe.

Invest : 3.500$
Mining : 2.800$ (with BTC at value 1000$!)

So if bitmine will not compensate this, every customer will make a big loss...;-(

the difficulty wont increase 20% over this year for sure.

Miners mine btc, not fiat.

Bullshit. You mine for profit. You sell your coins for fiat, gold, crackers, whatever. Both are valid metrics, and neither are mutually exclusive.

No asic manufacturer is going to guarantee your returns, as they cannot. If they say they are, RUN, Do Not Walk, away from that scam.

Also, there is something in the mind of a miner that loves the game, which is in itself a profit. Yes, historically and probably in the future, buying and holding will make MORE profit. But it's just not the same. Cheap beer will get you drunk, but it's not Fat Tire. Ya know?

[Wolke]

anybody who has a date of shipping ?