ASICMINER: Entering the Future of ASIC Mining by Inventing It

[DeathAndTaxes]

    So then my dream for ASICMINER can't come true   Thank sucks. There's no way to develop a chip that would share it's hash power with the operator and ASICMINER? If someone could figure out away to do this it would not only be good for the producer of them but it would be great for the network as a whole.
     Anyway thanks for the critique guys guess i gotta put more thought into it, i just don't get how else you franchise mining? How do you think he plans on keeping his franchisees honest and paying him for running the hardware?  

You are essentially talking about hardware level DRM.  Can it be done?  Sure.  However currently ASICs are relatively "stupid".  They have no idea what they are doing and that is a great thing.  You use "smart" and insanely cheap microprocessor to do the high level thinking and you connect them to banks of ASICs which are glorified calculators.  Yeah they are amazingly cool in terms of efficiency and speed but they are brain dead stupid.  They have no "concept" of what mining is.  They take a binary blob, double hash it and compare it to another binary blob to see if it is smaller.  All the "smarts" are in the miner software running on a cheap off the shelf microprocessor.

Moving all the mining logic into the ASIC would be very expensive for essentially no gain.  It could be done but the high level mining logic isn't a bottleneck so you aren't going to make it faster, and you certainly aren't going to do it cheaper then mass produced microprocessors running  open source code.  Still you would be able to "lock" the chip down ... until someone figures out a way to hack it and shares it with everyone on the internet who in mass disable the "mining tax".  All DRM to date has been broken, someone would figure out a way to break this too.   I mean people spend collectively millions of hours hacking video game consoles so they can play pirated games for cheap.  This is a box which makes money and hacking it would allow you to make MORE money.   If people are willing to devote so much time to hacking hardware to play cheap video games, how much time do you think they would devote to hacking hardware which makes free money?

[Rival]

With the price of shares of AM so closely tied to the btc/fiat rate, it is becoming difficult to calculate the correct ROI, and to determine if it is in fact positive.

At $90 per bitcoin, AM shares averaged around 4.5 each. 4.5 * 90 = $405.

At $120 per bitcoin, AM shares averaged around 3.3 each. 3.3 * 120 = $396.

At $130 per bitcoin, AM shares average around 2.9 each. 2.9 * 130 = $377.

Today:

At $140 per bitcoin, AM shares average around 2.5 each. 2.5 * 140 = $350.


This apparent loss of fiat purchasing power as the price of btc increases may be replaced by the dividends. They are only a few dollars at best per week, but  there are multiple dividends per price period so the accumulation of dividends might offset most of it. Of course, there are other factors such as other IPO, breaking news, FUD, etc. But a clear pattern is establishing itself. Could it be possible that, just like buying an asic (in most cases anymore), you are better off just sitting on your BTC than you are in investing? Help me out here and show me what I am missing.

[Jutarul]

Yeah, there will always be some ways to trick the system... That is why franchising works only within a "continued" relationship, with trust and/or collaterals involved.

And a "continued" relationship is actually desirable over a technical solution, which may be circumvented and cut ASICMINER off from the revenue stream.

One of the added values of the franchise business model is that it creates a culture of a sustained customer relationship, with a clear division of labor. The miner benefits from the reduced cost of scaling and reduced risk exposure to hardware depreciation. The supplier (ASICMINER) benefits from improvements in the cost structure of the mining facility, since competition will lead to franchisees competing for the deployed hash power by offering higher PPS (pay-per-share percentage = rental fees), causing inefficient miners to exit the market while keeping the revenue high.

The ability to push hash power into the market depends on the hardware production cost per PPS. That is where hardware suppliers will compete. The ability to run profitable depends on the upkeep costs per PPS. That is where miners need to compete. Hardware efficiency will be priced in by discounting the PPS correctly (inefficient hardware should trade at lower PPS). These principles of course are already at work right now, but the rather high hardware depreciation risk makes it hard to determine the expected profit from running mining equipment to a point where it effectively becomes a gamble. The franchise should allow both business partners to have less windfall and a more consistent revenue stream.

The first generation of ASIC miners are getting burned heavily. Other companies already have read the signs and are offering some form of income guarantee program (see, e.g. HASHFAST and their MPP). However, long term I suspect that only business models will succeed which have a guaranteed income stream for the miner at negligible risks. At least this should hold true if the majority of the miners are economical.

[Aedius]

With the price of shares of AM so closely tied to the btc/fiat rate, it is becoming difficult to calculate the correct ROI, and to determine if it is in fact positive.

At $90 per bitcoin, AM shares averaged around 4.5 each. 4.5 * 90 = $405.

At $120 per bitcoin, AM shares averaged around 3.3 each. 3.3 * 120 = $396.

At $130 per bitcoin, AM shares average around 2.9 each. 2.9 * 130 = $377.

Today:

At $140 per bitcoin, AM shares average around 2.5 each. 2.5 * 140 = $350.


This apparent loss of fiat purchasing power as the price of btc increases may be replaced by the dividends. They are only a few dollars at best per week, but  there are multiple dividends per price period so the accumulation of dividends might offset most of it. Of course, there are other factors such as other IPO, breaking news, FUD, etc. But a clear pattern is establishing itself. Could it be possible that, just like buying an asic (in most cases anymore), you are better off just sitting on your BTC than you are in investing? Help me out here and show me what I am missing.

Correlation does not equal causation. All else equal, investing in a BTC mining company is more of a leveraged bet on the price of BTC increasing rather than a hedge as some have described it. AMs share price has fallen due to a few different factors, including: competition appears to be increasing, total network hash rate has increased while AMs hashrate has stayed relatively flat, dividends have fallen slightly, and AM has taken some money out of dividends to fund future hardware.

IMO, AMs share price is VERY undervalued at current levels, Friedcat is one of the only mining operators who has consistently delivered if not over-delivered on his promises to the Bitcoin community and shareholders. AM has been consistently selling out of hardware exceptionally fast and will soon have next gen chips for mining and for sale, not to mention the new franchising business model which will allow AM to effectively have a greater percentage of the network than what would normally be possible for a typical mining operation.

Going to keep accumulating shares at these levels; enjoying the ~48% yield at current prices while the share price continues it's volatile course.

[ThickAsThieves]

Meanwhile, they are missing out on a lot of mining revenue.  We're now going on several weeks that the hash rate is stagnant (or dropping) and the network % is dropping.  We're talking thousands of btc of lost revenue in the last month.

Maybe they sold all the hardware instead of adding it to the farm. Maybe some hardware went to franchises. Maybe AM knows what they are doing.

Or, maybe the forums know better than AM?

So many arguments here operate on the premise that AM somehow chose to do something stupid regarding any given topic, why?

[Keyser Soze]

So many arguments here operate on the premise that AM somehow chose to do something stupid regarding any given topic, why?

People are accustomed to others involved in bitcoin making poor decisions (or outright scam).

[gog1]

With the price of shares of AM so closely tied to the btc/fiat rate, it is becoming difficult to calculate the correct ROI, and to determine if it is in fact positive.

At $90 per bitcoin, AM shares averaged around 4.5 each. 4.5 * 90 = $405.

At $120 per bitcoin, AM shares averaged around 3.3 each. 3.3 * 120 = $396.

At $130 per bitcoin, AM shares average around 2.9 each. 2.9 * 130 = $377.

Today:

At $140 per bitcoin, AM shares average around 2.5 each. 2.5 * 140 = $350.


This apparent loss of fiat purchasing power as the price of btc increases may be replaced by the dividends. They are only a few dollars at best per week, but  there are multiple dividends per price period so the accumulation of dividends might offset most of it. Of course, there are other factors such as other IPO, breaking news, FUD, etc. But a clear pattern is establishing itself. Could it be possible that, just like buying an asic (in most cases anymore), you are better off just sitting on your BTC than you are in investing? Help me out here and show me what I am missing.

The bitcoin / AM share price correlation theory again - which I highly doubt.  I think share prices slump because of the low dividend and apparent low share of network hash power.  Even if bitcoin is $45, would you be willing to pay 4.5 BTC for a share that pays out only 0.012 BTC / week?  NO.  Let alone the 9 BTC price base on your theory.

[velacreations]

Meanwhile, they are missing out on a lot of mining revenue.  We're now going on several weeks that the hash rate is stagnant (or dropping) and the network % is dropping.  We're talking thousands of btc of lost revenue in the last month.

Maybe they sold all the hardware instead of adding it to the farm. Maybe some hardware went to franchises. Maybe AM knows what they are doing.

Or, maybe the forums know better than AM?

So many arguments here operate on the premise that AM somehow chose to do something stupid regarding any given topic, why?

well, we're making assumptions using very little information.  There could be a lot of scenarios where they can't physically add more hashrate to the farm.  Or maybe there was a major flaw in the design of the new blades that has prevented their release.

In the end, we don't know "why", we just know the result: no hashrate increase and no new blades.

FC is likely making the best choices given the information he has.  I don't doubt he knows what he is doing, but that doesn't change the fact that AM is missing major revenue.

[ThickAsThieves]

I don't doubt he knows what he is doing, but that doesn't change the fact that AM is missing major revenue.

By this logic there are quite a few people & businesses missing major revenue though.

[velacreations]

I don't doubt he knows what he is doing, but that doesn't change the fact that AM is missing major revenue.

By this logic there are quite a few people & businesses missing major revenue though.

well, that brings us back to the original thing of "why"  Why can't AM take advantage of the clear opportunities at its feet? 

As we've seen in the last 4 weeks, AM clearly can't increase hashing rate, for whatever reason, nor release new hardware that has been promised since July.

I find that a bit worrying, to be honest.  Whatever obstacles that are preventing AM from moving forward may be significant obstacles that won't be resolved anytime soon.  Without more info from FC, there is no way to tell for sure. 

Are these obstacles something AM can't overcome?

[OneBTCJay]

I think patience is the biggest obstacle that AM investors need to overcome. People act like putting together a 50 THs farm is nothing more than pre-ordering 100 baby-jets.

AM wouldn't be here if they didn't have their act together. FC has been planing expansion from the beginning and has been doing great.

The bears have been loud recently and have used this two weeks of lower hash-rate to get the lowest share price, but I don't think it will last long.

[Vycid]

I don't doubt he knows what he is doing, but that doesn't change the fact that AM is missing major revenue.

By this logic there are quite a few people & businesses missing major revenue though.

Those businesses usually see a share price correction,  as is necessary and appropriate.

The correction that has occurred so far is not sufficient.

[velacreations]

I think patience is the biggest obstacle that AM investors need to overcome. People act like putting together a 50 THs farm is nothing more than pre-ordering 100 baby-jets.

well, we had that done some time ago.  It seems like getting a 50 TH/s farm up isn't too much of an issue for FC and team, but keeping it at 50TH/s and expanding beyond that is a significant hurdle.

FC has been planing expansion from the beginning and has been doing great.

up until about a month or so ago, and expansion has completely stagnated. 

The bears have been loud recently and have used this two weeks of lower hash-rate to get the lowest share price, but I don't think it will last long.

based on what?  I don't see any signs of the mining operation improving, and hardware delivery has been stalled waiting on the new blades. 

[DeathAndTaxes]

I think patience is the biggest obstacle that AM investors need to overcome. People act like putting together a 50 THs farm is nothing more than pre-ordering 100 baby-jets.

well, we had that done some time ago.  It seems like getting a 50 TH/s farm up isn't too much of an issue for FC and team, but keeping it at 50TH/s and expanding beyond that is a significant hurdle.

FC has been planing expansion from the beginning and has been doing great.

up until about a month or so ago, and expansion has completely stagnated.  

The bears have been loud recently and have used this two weeks of lower hash-rate to get the lowest share price, but I don't think it will last long.

based on what?  I don't see any signs of the mining operation improving, and hardware delivery has been stalled waiting on the new blades.  

Has anyone considered what 65 TH/s of 130nm tech uses in power?  65 TH/s @ 8.5 J/GH is ~ 550KW or >2200 Amps @ 240V.    They have enough floor space, power and cooling to support >550KW of equipment.  If they are planning to migrate to more efficient "next gen" tech and they are near capacity why pay for the cost to expand if they don't need to. To illustrate say AM's 2nd gen chips are ~0.6 J/GH at the wall.  The existing space, power, and cooling could support not 65 TH/s but >900 TH/s.  If FC knows superior tech is coming and they are close to their limit on power that their current buildings can handle why would they want to expand to additional buildings rather than sell the excess rigs at what I consider to be a very good price.   The selling price isn't leaving that much net profit on the table.  So while you wait for 2nd gen chips, maintain the existing farm and sell excess capacity to improve cashflow.  When the 2nd gen blades are ready, replace the existing farm with current one, sell of the old tech and expand the farm to 1 PH/s.

It is very possible that AM will NOT expand their hashrate to more than 50 to 60 TH/s until they can replace existing tech with next gen tech.


on edit: edited for brevity and clarity.

[Ninshatamoto]

I think patience is the biggest obstacle that AM investors need to overcome. People act like putting together a 50 THs farm is nothing more than pre-ordering 100 baby-jets.

well, we had that done some time ago.  It seems like getting a 50 TH/s farm up isn't too much of an issue for FC and team, but keeping it at 50TH/s and expanding beyond that is a significant hurdle.

FC has been planing expansion from the beginning and has been doing great.

up until about a month or so ago, and expansion has completely stagnated. 

The bears have been loud recently and have used this two weeks of lower hash-rate to get the lowest share price, but I don't think it will last long.

based on what?  I don't see any signs of the mining operation improving, and hardware delivery has been stalled waiting on the new blades. 

Woah woah woah.....Velacreations is going bearish on AM?!

I really need to reconsider my position.

[binaryFate]

I think patience is the biggest obstacle that AM investors need to overcome. People act like putting together a 50 THs farm is nothing more than pre-ordering 100 baby-jets.

well, we had that done some time ago.  It seems like getting a 50 TH/s farm up isn't too much of an issue for FC and team, but keeping it at 50TH/s and expanding beyond that is a significant hurdle.

FC has been planing expansion from the beginning and has been doing great.

up until about a month or so ago, and expansion has completely stagnated. 

The bears have been loud recently and have used this two weeks of lower hash-rate to get the lowest share price, but I don't think it will last long.

based on what?  I don't see any signs of the mining operation improving, and hardware delivery has been stalled waiting on the new blades. 

Has anyone considered what 60 TH/s of 130nm tech uses in power?  If FC knows superior tech is coming and they are close to their limit on power why lease an entire second building to build out massively using 130nm tech which will be obsolete soon.  Just maintain the network, sell off the excess and prepare to move to a more efficient processing node.

65 TH/s @ 8.5 J/GH is ~ 550KW or >2200 Amps @ 240V.  The existing hashrate provides insight into the fact that AM has facilities which can handle 550KW of direct power and more importantly the ability to remove 550KW of heat (which likely means up to another >200KW of AC).  Say AM's 2nd gen chips are ~0.6 J/GH at the wall.  The same amount of facilities could handle > 900 TH/s of capacity and that is just an estimate based on know hashrate.  I doubt they are running to the last milliamp so the goal of 1 PH/s makes sense. 

However to expand significantly before the next gen chips arrive may require additional facilities.  Have you looked at the selling prices AM gets for their gear.  I mean by my math it is pretty close to the NPV of the lifetime revenue stream based on difficulty growth.  AM isn't leaving that much on the table by selling the hardware rather than self mining.  If they are near capacity (for 130nm) why spend money on more facilities you don't need to get to 1 PH/s and mine when you can get almost the same net revenue by selling the gear to someone else?

Interesting hypothesis, this sounds plausible indeed. But gen2 is not expected before November, that would imply we will stay at 60TH/s until November/December, and hardware sells will need to be enormous to compensate the lack of mining dividends. 60TH/s in 2 months will be nothing, maybe 2%.

[DeathAndTaxes]

Interesting hypothesis, this sounds plausible indeed. But gen2 is not expected before November, that would imply we will stay at 60TH/s until November/December, and hardware sells will need to be enormous to compensate the lack of mining dividends. 60TH/s in 2 months will be nothing, maybe 2%.

Agreed and maybe they have some excess capacity to expand marginally between now and then  but I don't see AM hashrate going to say 120 TH/s using 130nm tech.  That would mean doubling power to over a megawatt.  One other factor is that "something" happened in the farm over the last month.  Looking at the 3 day average ( http://asicminercharts.com/ ) on 8/23 the 3 day average rate peaked at 53 TH/s and then over the next 4 days declined to half that.   Now say you are FC and you have 60 TH/s of hardware deployed but are running into problems getting 60 TH/s out of your 60 TH/s of hardware and another 10TH/s of boards arrive does it really make any sense to install them and possibly compound your problems.   AM Blades somehow still sell very close to their expected net revenue so just sell the hardware (for roughly the same amount as you would get mining from them over the next couple months) while you can still get a good price.  

Having run a 10KW GPU farm for over a year, if FriedCat can successfully manage 50x the power AND has a plan to more efficient tech, well that is worth more than a stable dividend in the long run.  I think many of these other mining ops have no idea the complications that "real" power loads bring to the table.    Going from 200W to 1K of mining gear is pretty simple just plug it in and go; going from 100KW to 500KW isn't quite the same thing.   I would like to see the hashrate stabilize north of 50 TH/s for more than a week and some communication on the plan to migrate to 2nd gen.

[Ytterbium]

  If people are willing to devote so much time to hacking hardware to play cheap video games, how much time do you think they would devote to hacking hardware which makes free money?

Because those systems weren't all integrated onto a single die. Right now the tech to 'hack' a single chip doesn't really exist (or at least, it costs more then just making your own chips)

You could do a DRM miner, essentially you'd just have it check a digital signature on the block header, and require it to do 20 signed work units for every unsigned one (or whatever).

But the problem is you'd need to add a lot of transistors, more then the number required to simply hash the block header.

[Ytterbium]

Having run a 10KW GPU farm if FriedCat can manage 50x the power AND has a plan to more efficient tech then that is worth more to me than a stable dividend in the long run.  I think many of these other mining ops have no idea the complications that "real" power bring.    Going from 200W to 1K of mining gear is pretty simple just plug it in and go.  Going from 100KW to 500KW isn't quite the same thing.   I would like to see two things though.  The first is that the hashrate stabilize north of 50 TH/s for more than a week and some communication on the plan to migrate to 2nd gen.

Freidcat can probably afford his own coal plant at this point. (Coal being the go-to fuel source in China at the moment.  In the US natural gas is now more popular)

[binaryFate]

Agreed and maybe they have some excess capacity to expand marginally between now and then  but I don't see AM hashrate going to say 120 TH/s using 130nm tech.  That would mean doubling power to over a megawatt.  One other factor is that "something" happened in the farm over the last month.  Looking at the 3 day average ( http://asicminercharts.com/ ) on 8/23 the 3 day average rate peaked at 53 TH/s and then over the next 4 days declined to half that.   Now say you are FC and you have 60 TH/s of hardware deployed but are running into problems getting 60 TH/s out of your 60 TH/s of hardware and another 10TH/s of boards arrive does it really make any sense to install them and possibly compound your problems.   AM Blades somehow still sell very close to their expected net revenue so just sell the hardware (for roughly the same amount as you would get mining from them over the next couple months) while you can still get a good price. 

About this "something" that happened this month, when the hash rate was low, it was about half the "normal" one. Maybe they were moving to a different location, moving half of the hardware at a time? It would be possible that the initial location does not fit the needs for further expansion, be it in terms of space or electricity, or even safety or whatever may not be super urgent at the beginning of the operation but becomes relevant when the company consolidates.

I would like to see two things though.  The first is that the hashrate stabilize north of 50 TH/s for more than a week and some communication on the plan to migrate to 2nd gen.

I doubt that we will get much communication about 2nd gen at all. This would be very valuable to competitors, actually I prefer not to have this information... If friedcat does it purely "friedcat's style", we will get an unexpected message one day that announces some awsome specs and that gen2 hardware is in stock ready to ship.