Topic: CoinTerra announces its first ASIC - Hash-Rate greater than 500 GH/s - page 110. (Read 231016 times)

Also, people will keep mining when it's unprofitable if they are bullish on bitcoin's price... i.e. everyone who purchased mining hardware. I'm pretty sure that most of us wouldn't be here if we didn't think that the $1.2 billion that BTC currently represent will go way up as adoption continues to grow, taking the value of each individual bitcoin with it.

Even if someone were to do the math, it wouldn't be accurate as everyone is Pre-Ordering and will end up holding the bag well after it has stopped being profitable.  If you could only buy immediately available hardware, then an accurate assessment could be made.

Excellent point, but do me one favor, someone who's happy to do math right now, what difficulty is justified by a $110 price? If 1 trillion diff corresponds to $42k price for BTC, what BTC price of $110 corresponds to which diff?

That should give us the end-point of difficulty leveling off and should also tell us roughly when to expect that to finish levelling.

a) KNCminer (price reduction either in November or for orders for November, can't remember)

b) XCrowd  as stated in https://www.xcrowd.co.uk/mining/home.html

Don't bet your ass!

This fit KNCminner

The ASIC vendor that first does this with stable supply will be the one that survives. We now have multiple 28nm designs coming out that are probably close to the most optimal design, with only marginal improvements going forward.

Once an ASIC vendor has sunk the NRE costs and shipped 1 unit, each additional unit is quite cheap to produce. In the end it will be a race to the bottom price wise for ASIC vendors trying to recoup their NRE funds, or maximize profit. Those that are paying a price premium today for hardware will see that they paid 10x what those units cost 6 months later.

In the end it will settle down similarly to the FPGA era, where the hashrate stabilized for 1 year and individuals could buy new hardware expecting a 2 year ROI in a flat BTC environment (with bonus profitability during upward price movements).

What company do you mean?

Alright, I'll say it.

Alot of people are going to lose their pre-order money.

Why?

Because these companies (almost all of them) are acting as if they aren't shipping out .4 Terrahash per device.

I think only 1 (ONE) company has even acknowledged publicly that they will incrementally lower their prices as market conditions worsen.

I think only 1 company has acknowledged that they have a plan to compensate buyers with chips if they can't make ROI. But only chips.

-----------------------

I have yet to see any company acknowledge that they need to start work on large parallelism of their current designs. Not 1 so far....

You'll probably wonder why this matters. The reason why it matters is because most have already reached the lowest nm process they could possibly ever afford. And now, they only have two options. Make design changes (at the chip level) and re-charge all their customers for future chips at the same nm level.

Or
(the most likely IMO)

They will have to figure out a way to change their designed platforms and prices to adapt to massively parallel operation. A person only has so many plugs in their home. That is a limiting factor. There is also the problem of heat dissipation spread out over one room. There is a wall coming in the not so distant future where they are going to have to start thinking outside the box to create power efficiency.

If they don't their company is effectively dead.

If they aren't thinking about it before their first units come online, then you can pretty much expect that the company will fail at some point when they can't lower their prices due to base hardware costs and miners can't get a decent amount of profit they really desire in the next offerings.

I am not interested in two months into the future.   I am mostly interested in 2014 EOY.   Think about where the HR can go before rational people stop buying.   And where it can go because rational people do not know how many TH/s has been pre-ordered and will be delivered (nor when).
The one thing that changes the slope of the curve is the END OF PRE-ORDERS and the first company that starts shipping same day as ordered, changes a lot.   Think about what I mean.

Cointerra is taking happy pills and making things look like it is fine and their machines are truly profitable for more than 2 months.

They say this with the full knowledge that the other companies surrounding them are about to ship many thousands of individual units in increments of 400+ GH/s.

Think about that. Stop being a pussy and stop avoiding reality folks. If you think the current curve is steep, wait till later.

========================

The good news is that ASICs must become cheaper and I think all the large companies behind closed doors know this. They just don't want to acknowledge this in open public view. Otherwise their pre-orders will significantly downsize prematurely.

---------

Are you placing your pre-order centered on the reality at the time of delivery or centered around a fantasy of the present being applicable to the future where each customer can turn on _significant multiples_ of _A Terrahash_ that then are coming online?

The only thing I see happening is that CoinTerra is not lowering it's prices. It's set them pretty high.

There was a unanimous opinion that it is set too high and is rediculous. Will some pay it? Yes, then they will lose money to basic math and reality.

----------------------

The rest of the miners who want to believe in a pipe dream are crying out loud in public "It can't be so!". The math must be wrong? Bullshit.

You folks just don't want the companies to come to the conclusion that they either need a game plan that accounts for these violent market changes or they will be out of business before they even ship.

What happened to Avalon will happen again, only to multiple companies if they are not prepared. Reality has already rung your doorbell. You can either ignore it (and go out of business) or adapt.

Funny, my results lined up 2 months into the future within an error margin of just a few days.

Can I say that this will be true in 6 months? No, why? Because reality is showing an even steeper curve than I used in Excel. So my models will at best, a best case scenario.

I do not understand why you bother with the sarcasm.   It will be lost on them.   No one is going to read the long string that you put up that had thoughtful analysis of where the network could go based on marginal costs and highlighting how the pre-order process effs up rational economic decision making.   Stop wasting your typing.   Let them use their website models to explain everything to themselves.     A great question at HBS in first year finance for modeling "If you model a company growing at 6% per year and USA GDP is growing at 3%, how large that company become over time?"
Excel has allowed people with no understanding of compounding to come up with solutions that do not pass the most basic sanity checks.

You got some good points here.

I expect that few of ASIC manufacturers that hoped to earn a lot of cash on "selling shovels" will start to mine instead of selling anything to the public. It will be their (desperate) attempt to recover NRE costs. They could build dedicated facilities in somewhere where electricity is dirt cheap and put all small-time-miners out of business thanks to economic of scale (28nm silicon cost should be around 1-2$/GH excluding NRE).

Centralization of Bitcoin mining might happen sooner than anyone could have expected...

You post pretty much sums up this company.

Everyone should read it at least once.

I have excepted a lower price than 15K for 2TH/s as well for the point in time of delivery and my interpretation of their promises.
In addition I can not afford to spend that much which increases my disappointment.

@cointerra: I'm looking forward for lower priced products (starting at $500)

Anyways I'm a little confused about the negative ROI predictions mentioned here several times.
When I put in the following numbers as start parameters which is not likely to be happen at all I assume:

2TH/s,
1200W,
hw=$15750,
btc=$100,
electricity=$0.15
1.000M difficulty with 5% increase

=> It shows to me 33% ROI in the first year and break even after 250days when btc is $100
=> Even at difficulty 2.079M monthly revenue is $1,420 and profit $1,290
http://mining.thegenesisblock.com/a/569aadb633

At 10.000M difficulty you still make >$160 a month,
at a difficulty of 22.800M and BTC=$100 you reach the $0 profit line: http://mining.thegenesisblock.com/a/137e040147

So I do not understand that it will never be possible to get ROI in fiat - am I missing a major point 

Do you just complain for the fun of it?
When you do your calculations with difficulty > XXXXM, do you even think about what 10% increase means..?
Do you expect a money printing machine paying your daily life ??
Serious questions I'm asking myself..

You have to realize that Cointerra got into this as blind as customers of other ASIC vendors. Cointerra couldn't predict Hashfast nor KNC nor Bitfury, etc., etc. Matching price is a necessity forced on them that is also a big bummer for them. They were likely hoping on a price at least 3 times that amount.

You figure the cost of this custom hardware after NRE probably comes down to a few hundreds of dollars.

There was major major major money to be made, and Avalon proved it charging $7k for some ASIC hardware that was otherwise inexpensive to make. Avalon probably made 500% or more on each unit, who knows.

Huge profit margins draw competitors. They've arrived now. But they're not able to charge what they were hoping they'd be able to charge. They're just as aced out as the people who bought miners looking at what they could've generated months ago (like Avalons were 15 bitcoin a day back in april).

Price competition has not only hit, it's going to scare off other competitors. The value of hardware is going to crash so quickly after september that it's likely that many people formerly planning to startup an ASIC venture have now tossed those plans entirely, too much competition. And now people are worried that Nov/Dec asics might not ROI at all.

Mining should always return to the point where the cost of electricity is just under the cost of mining bitcoin, long term.

If I were an ASIC company wanting to stay in this biz long term rather than going for the homerun, facing all these fence-swingers right now, I'd be making incredibly high efficiency ASICs, air cooled, and plan to sell them for a slight markup and make up the difference on volume. A 10gh USB stick for $20 would be a market winner, even as the large asics begin dying like flies at the end of this year.

+1