ANTMINER S7 is available at bitmaintech.com with 4.86TH/s, 0.25J/GH

[carlosmnk]

ANy one have S7 coupon sell ? PM thanks.

Yes, there are some people selling them, but not here. Try at "Bitcoin Forum > Economy > Marketplace > Goods > Computer hardware": https://bitcointalk.org/index.php?board=75.0

[1713860478]

1713860478

[1713860478]

1713860478

[1713860478]

1713860478

[1713860478]

1713860478

[Mr.paul]

ANy one have S7 coupon sell ? PM thanks.

Yes, there are some people selling them, but not here. Try at "Bitcoin Forum > Economy > Marketplace > Goods > Computer hardware": https://bitcointalk.org/index.php?board=75.0

Okie..thanks you

[lucky1974]

Based on what I saw, I'd estimate Bitmain was replaceing S5's with S7s for 2 months at most, and possibly with S5+ for a month or two before that (and then likely sold those S5+ as "new" units once they got the S7 into production).

 I don't believe for a microsecond that the S7 has been in production for 6+ months.

yea okay and you own bitmain right.

[notlist3d]

Confucius says: Man that needs loan to buy bitcoin miners, has hole in pocket

Very wise

I would agree.  I only invest my own money in mining.   I can't imagine taking a loan for a miner. 

So much is unknown just like difficulty.  I can't guarantee I will ROI so how can I risk others money on that same investment.  I do it out of a hobby I love, and a hope of ROI.

Confucius says: Man that uses his own money to buy bitcoin miners, has overdraft bank account.

Confucius is wrong.  Confucius should only spend some of his money on a hobby not all

[notlist3d]

Based on what I saw, I'd estimate Bitmain was replaceing S5's with S7s for 2 months at most, and possibly with S5+ for a month or two before that (and then likely sold those S5+ as "new" units once they got the S7 into production).

 I don't believe for a microsecond that the S7 has been in production for 6+ months.

When you say what you saw what are you referring to?  Did you get to see a plant? Just hashing online?   What specifically?

[Finksy]

you just need to handshake with the top 3 or 4 builders to increase as a team.  

I'll go ahead and stop you right there Phillip    Have you ever seen anything that has lead you to believe there is even professional courtesy for one another among the big dogs, let alone actual collusion?  That is a massive assumption, and one that I do not believe to be true. There will always be the company that thinks that they will be better, more efficient, more dense & finished sooner, and would only be helping their competitors by putting a leash on their own hashrate.

It's the prisoner's dilemma, that which is best for all as a whole will often not be the decision made by the individuals.  Network hashrate has been increasing almost continuously, and it takes many more PH/s at every adjustment to maintain your chunk of the revenue stream.  If what you said was true, what incentive would they have for developing new miners as quickly as possible with higher density?  I would bet that the amount of money saved on electricity over an entire year (if in-house hashrate were to remain flat) would not make up the amount of money required to develop and build a new generation chip & miner.  They do it because A) they feel they need to push it further in order to survive and B) they want to get an edge on competitors and drive the smaller ones out of business, freeing up more revenue for themselves.  Consider this, if they had kept hashrate constant over the past year by colluding, AMT/BFL (among other bankruptee's) hardware would still be profitable and competing with them on the network today.  By driving the hashrate higher, they knock more people out of the runnings, and take a higher percentage for themselves (even if it comes at a cost).

I do agree with you that if one of the big builders really believed they had designed  a big edge they would try to get more hashrate.

And there is the truth Phil, that's what they all believe.  Look at the bullshit PR and fluff that goes on by the manufacturers (look at Bitfury's Twitter page for some examples).  It's all secrecy and speculation, to try to discourage your competitors from further development.  (read: the exact opposite of collusion)

If you believe that KNC, Bitfury, Innosilicon, SP-Tech and Bitmain have a handshake agreement to keep the network flat, I'm not sure what to tell you except look at the fact that the network hashrate has doubled in the past year (which has been mild compared to the prior year). The reason it has been increasingly more slowly recently is due to the fact that many of the competitors have been knocked out of the game (due to the fact that they keep driving it higher), or were in between generations.

[notlist3d]

you just need to handshake with the top 3 or 4 builders to increase as a team.  

I'll go ahead and stop you right there Phillip    Have you ever seen anything that has lead you to believe there is even professional courtesy for one another among the big dogs, let alone actual collusion?  

They can work together when they need to.  Look at the 51 percent day's with ghash.  Do you honestly believe the companies were not talking to each other for a fix and trying to decide what to do?

Chances are they did not do this on the S7 as far as making it.  I doubt they told other companies they would keep 28 NM and let other companies spend much more time and money on lower nm.   So they are also able to keep secrets to.

[Biodom]

you just need to handshake with the top 3 or 4 builders to increase as a team.  

I'll go ahead and stop you right there Phillip    Have you ever seen anything that has lead you to believe there is even professional courtesy for one another among the big dogs, let alone actual collusion?  That is a massive assumption, and one that I do not believe to be true. There will always be the company that thinks that they will be better, more efficient, more dense & finished sooner, and would only be helping their competitors by putting a leash on their own hashrate.

It's the prisoner's dilemma, that which is best for all as a whole will often not be the decision made by the individuals.  Network hashrate has been increasing almost continuously, and it takes many more PH/s at every adjustment to maintain your chunk of the revenue stream.  If what you said was true, what incentive would they have for developing new miners as quickly as possible with higher density?  I would bet that the amount of money saved on electricity over an entire year (if in-house hashrate were to remain flat) would not make up the amount of money required to develop and build a new generation chip & miner.  They do it because A) they feel they need to push it further in order to survive and B) they want to get an edge on competitors and drive the smaller ones out of business, freeing up more revenue from themselves.  Consider this, if they had kept hashrate constant over the past year by colluding, AMT/BFL hardware would still be profitable and competing with them on the network today.  By driving the hashrate higher, they knock more people out of the runnings, and take a higher percentage for themselves (even if it comes at a cost).

I do agree with you that if one of the big builders really believed they had designed  a big edge they would try to get more hashrate.

And there is the truth Phil, that's what they all believe.  Look at the bullshit PR and fluff that goes on by the manufacturers (look at Bitfury's Twitter page for some examples).  It's all secrecy and speculation, to try to discourage your competitors from further development.  (read: the exact opposite of collusion)

If you believe that KNC, Bitfury, Innosilicon, SP-Tech and Bitmain have a handshake agreement to keep the network flat, I'm not sure what to tell you except look at the fact that the network hashrate has doubled in the past year (which has been mild compared to the prior year). The reason it has been increasingly more slowly recently is due to the fact that many of the competitors have been knocked out of the game (due to the fact that they keep driving it higher), or were in between generations.

i think collusion is a strong word, but i also have a feel that hash rate is being "managed"  through periodic large % switching on and off-you can see it in graphs.
most likely large miners analyze the hash rate and see when it is the most beneficial to switch a largish % off (in the last few days before adjustment, of course) and when to switch it back on-in the beginning of the period, of course. Or, maybe nobody does anything proactive, and it is just statistical fluctuations.

[Finksy]

Re-reading my post, it comes off a little strong, I'm not trying to under-rate anyone else's views. All I'm stating is my opinion from a business point. However as we all know, Bitcoin is very different from most business ventures, so anything is possible.

[lucky1974]

GREED !

[philipma1957]

Re-reading my post, it comes off a little strong, I'm not trying to under-rate anyone else's views. All I'm stating is my opinion from a business point. However as we all know, Bitcoin is very different from most business ventures, so anything is possible.

It is hard to write the reasons i have for believing the top builders manage the rate of growth. but here goes.

100% = the hashrate.

15% bitmain

15% bitfury


15% sptech

15% knc

15% avalon

15% 21? can not remember name

10% us little miners farms

if any builder builds a ton of new gear all other builders see it and build more gear to catch up.

so new gear is limited in build up.

wheter they handshake talk on the phone or just look at hashrates does not matter it is the econmic nessity that creates paced growth.

so new gear ie new s-5 is not desired that much.

so growth has long slow rates when new gear is the same as last weeks gear.

Better gear  like the s-7 does much more upward movement in diff.

power has meaning two cent power allows for new cheap gear or used cheap gear compete against the s-7.

so if bitmain has four cent power and an s 7

my two cent power in a spot that needs heat makes s 3

still earn money.



So  while the old gear shifts to cheaper spot places.

Bitmain can do better by swapping s-5 with s-7  not filling up all its power cap.

 IE 60ph  in = 90ph in s-7  with  30ph not built

or 60 ph in s-5  = 80 ph in s-7 with 40ph not built

or 60 ph in s-5 = 70ph in s-7 with 50 ph not built

or 60 ph in s-5 = 60ph in s-7 with 60 ph not built.

every case above is better then 60ph in s-5  with the s-7 not in existence.



this is true with any builder/miner   they all get this luxury of swapping slowly .

This is an economic realty and is why hashrate has not gone nots in the last 360 days.

Now if I invent (fat chance)  a 0.01 watt chip  or a 0.001 watt chip  put  it on a usb stick  boom we get a new philasic  chip race.

that is not happening.  what is happening is a slower deliberate swapping of gear.


 bitmaintech may end up doing 60ph s-5 to  120 ph s-7  but they only need do

60 ph s-5 to 60 ph s-7 to be far better off.

[dmwardjr]

Sounds logical to me.

[QuintLeo]

Based on what I saw, I'd estimate Bitmain was replaceing S5's with S7s for 2 months at most, and possibly with S5+ for a month or two before that (and then likely sold those S5+ as "new" units once they got the S7 into production).

 I don't believe for a microsecond that the S7 has been in production for 6+ months.

When you say what you saw what are you referring to?  Did you get to see a plant? Just hashing online?   What specifically?

 "Used" S5 sale.
 Hashrate fairly stable for the first part of that.
 ETC.


 If the S7 had been in production for 6+ months, there would have been a much longer and bigger hashrate bump due to Bitmain using them internally.

It is hard to write the reasons i have for believing the top builders manage the rate of growth. but here g

 It would be counterproductive for "the top builders" to manage the rate of growth, they'd be eating THEIR OWN PROFITS to do so.
 Consider "market share" and how effective it can be in the long term to dominate market share.
 Additionally, there are limits to how much gear any builder CAN build, none of them have huge financial resources.

 The REAL reasons for slower growth rate of late are
(1) the bitcoin price collapse, giving much less incentive to folks to mine and lower profits for those that DO mine and sellers of mining gear
(2) the tech race is slowing a TON as the tech used to mine Bitcoin gets close to the best AVAILABLE tech
(3) the COST of developing new tech gets a LOT higher as you get close to the "bleeding edge" on tech
(4) the death of some miner companies due to the above factors making them unprofitable

 There is ZERO "economic need" to pace growth rate, in fact there is at least one good reason to NOT pace growth (ref market share) for any single company, and there have been way too many cases in history where it's hasn't happened even where it was a GOOD IDEA to do so (look at the history of Oil Production for a well-documented example).



 Do keep in mind that there is a significant random factor in hashrates.

[notlist3d]

Based on what I saw, I'd estimate Bitmain was replaceing S5's with S7s for 2 months at most, and possibly with S5+ for a month or two before that (and then likely sold those S5+ as "new" units once they got the S7 into production).

 I don't believe for a microsecond that the S7 has been in production for 6+ months.

When you say what you saw what are you referring to?  Did you get to see a plant? Just hashing online?   What specifically?

 "Used" S5 sale.
 Hashrate fairly stable for the first part of that.
 ETC.


 If the S7 had been in production for 6+ months, there would have been a much longer and bigger hashrate bump due to Bitmain using them internally.

I would say before they sold the used S5's they likely had the chip design ready for S7's and likely working prototypes.   I thought you had a inside source or saw something we didn't the way you worded you saw something.   

It's really speculating not "seeing" i think that you are doing.   Just as I am doing.

[tonycamp]

really guys S7 S5 prob both good i dont know prices but you miners of BTC must be aware of this prices and caculations for the all mining profitblility and prob thers also second hand miners that are also good for reasenable prices if this its not top of cherry cake prob you dont need the top just a piece of the miners that also its better than cloud mining that promisses high rois and the best for a start prob its a second hand prospero or sometning nothing like this s7 s5+ that are soo expencive and to get into investment can be very damn uncarfull

[Searing]

Re-reading my post, it comes off a little strong, I'm not trying to under-rate anyone else's views. All I'm stating is my opinion from a business point. However as we all know, Bitcoin is very different from most business ventures, so anything is possible.

It is hard to write the reasons i have for believing the top builders manage the rate of growth. but here goes.

100% = the hashrate.

15% bitmain

15% bitfury


15% sptech

15% knc

15% avalon

15% 21? can not remember name

10% us little miners farms

if any builder builds a ton of new gear all other builders see it and build more gear to catch up.

so new gear is limited in build up.

wheter they handshake talk on the phone or just look at hashrates does not matter it is the econmic nessity that creates paced growth.

so new gear ie new s-5 is not desired that much.

so growth has long slow rates when new gear is the same as last weeks gear.

Better gear  like the s-7 does much more upward movement in diff.

power has meaning two cent power allows for new cheap gear or used cheap gear compete against the s-7.

so if bitmain has four cent power and an s 7

my two cent power in a spot that needs heat makes s 3

still earn money.



So  while the old gear shifts to cheaper spot places.

Bitmain can do better by swapping s-5 with s-7  not filling up all its power cap.

 IE 60ph  in = 90ph in s-7  with  30ph not built

or 60 ph in s-5  = 80 ph in s-7 with 40ph not built

or 60 ph in s-5 = 70ph in s-7 with 50 ph not built

or 60 ph in s-5 = 60ph in s-7 with 60 ph not built.

every case above is better then 60ph in s-5  with the s-7 not in existence.



this is true with any builder/miner   they all get this luxury of swapping slowly .

This is an economic realty and is why hashrate has not gone nots in the last 360 days.

Now if I invent (fat chance)  a 0.01 watt chip  or a 0.001 watt chip  put  it on a usb stick  boom we get a new philasic  chip race.

that is not happening.  what is happening is a slower deliberate swapping of gear.


 bitmaintech may end up doing 60ph s-5 to  120 ph s-7  but they only need do

60 ph s-5 to 60 ph s-7 to be far better off.

WARNING: I got on a roll below...if you don't like me put me on /ignore....RANT below

On the other side of the fence with scrypt miner gear and LTC in particular I've managed to run my Titan(s) way way longer mining LTC for the fact that when LTC price was
low at the beginning of the year MOST if not all the pow/scrypt BIG miners like 300mh and above were canceled.(volcano etc)....also KNC seems to have hyped its 2nd batch and beyond
units as being a lot more then actually shipped and canceled their 150mh last set of miners too boot thus in this universe you had the titans on the big miner end and
everyone else with no new miners coming out of any large number spring summer and fall ..ie thats all there is folks (at this point in time anyway)

thus the result was (w/o collusion) a more or less FLAT network where I mined the frack out of stuff as others fell off the pow/scrypt miner bandwagon from around feb till
the LTC halving due to older miners electric etc

so as a result i got to thinking along the same lines...IF (this is horrible future guess).....it got to the point where KNC got to the place they say they want to be 30% of the
network and 16nm chips and you have say another few for a total of say 5 asic chip manufacturers who plan to use as they seem to say like knc all their 16nm or 14nm chips for their own private IPO $$$ data halls...(for example knc/bitfury/sfards/innosilicon/avalon? anyway figure 5 of them all got the chunk they want and 80% plus of all mining and they have the
newest baddest next gen chips for their own use).....well I mean really think along how these guys have shown their stripes in the past

now you and the others soon are chugging away with their 16nm 14nm data halls merrily away....and you ANNOUNCE you are gonna make the next generation jump in
chip technology..at least in DESIGN (or so you say) .beyond 16nm ..but hold off watching your competitors to see if they jump first.......they in turn are doing the same have say a design plan in place but are waiting to
pull the trigger...and everyone sits on the 'nuclear option' as long as possible
..this is not conclusion per say...BUT IMHO there will be very little motivation to pull that trigger if they all can avoid it ....and with the costs/risks/$$$ involved I'd attempt to hold back and compete with them with them (at this point in time
with probably this flatter network rise) by simply using the gen chip tech i have for example say 16nm knc to say....and just hedge with my competitors with a NEW BIGGER data hall of the
same generation chips..they in turn would do the same...keep the chips you got ...look at competitors .....diff creeps up modestly from its flatline direction and expand or put in
another data hall in a round robin way to compensate as your competitors do the same..a mexican standoff kind of collusion of the network assets you all control and the difficulty rise.

I'd compete with expansion and bigger badder data halls with the gen chips I have now with my newbie IPO $$$ and cheap electricity as a sure bet as long as I thought I could
get away with it and probably everyone else would do the same


BUT NO ONE would want to nor have a reason to start the arms race of the next chip generation under that scenario ..if they could 1) keep the network someone flatter
and compete with others with data hall builds and the round robin rotation of such and 2) with 5 of these companies should my scenario be true having above say 80% or more
of the mining of bitcoin you could just BET ON IT there would be a huge push and pressure for making transaction fees MANDATORY and at a level that benefits them even more
without the need of further innovation on chips ..ie they have their cake and eat it too

this is very true in the world of oil for example ..no real collusion of oil companies per say ...but they sure are not building NEW refinery's as an example in the cost is too high
they simply are expanding the ones they have already due to the fact their are ..on the same theme...5 or so major oil companies ie why rock the boat

it just makes sense

NOW the point on will they do so ..not sure ...the ego of these guys thinking they can make the next gen chip with IPO newbie cash would be strong and try to bury the
completion with a great leap forward with your new chip...knc tried this imho but the thing  was taped out feb 2nd and they obviously did not get it out on the time line
they felt they should if it is deploying now (to the best of my knowledge that is) and if you miss the deadline ..the chip don't work etc and looking at your competiors deploying
the same new chips to further the arms race too boot thus your 1 year down the road plan you see the 4x the eff of your chip say corresponding in difficulty  you will have
to compensate for ...well hell...........slow things down to a walking pace...use data halls to compete on this flat network trajectory and lobby that mandatory transaction fees
are the answer is what I would do for an evil genius plan

(funny how much I think of evil genius plans since 2013 and knc went from nun to porn star because their was more money in being slimy)

anyway something to consider

my last point that could blow all this out of the water and all the above asic's into the dust bin is if someone like INTEL and/or AMD committed to an asic scrpt/pow btc and/or
LTC type chip because of their scale if they drank the kool aid they could eat these current asic companies up like m&m's on a 10 min work break imho would be quite the show
if INTEL drank the btc kool aid and made a chip let me tell you

but that is another tale for our crypto stories of unicorns/rainbows and fairy dust ..from drinking the crypto coin kool aid

end of rant

[fullzero]

The REAL reasons for slower growth rate of late are
(1) the bitcoin price collapse, giving much less incentive to folks to mine and lower profits for those that DO mine and sellers of mining gear
(2) the tech race is slowing a TON as the tech used to mine Bitcoin gets close to the best AVAILABLE tech
(3) the COST of developing new tech gets a LOT higher as you get close to the "bleeding edge" on tech
(4) the death of some miner companies due to the above factors making them unprofitable

 There is ZERO "economic need" to pace growth rate, in fact there is at least one good reason to NOT pace growth (ref market share) for any single company, and there have been way too many cases in history where it's hasn't happened even where it was a GOOD IDEA to do so (look at the history of Oil Production for a well-documented example).

 Do keep in mind that there is a significant random factor in hashrates.

Strange that you now seem to have the same opinions you argued agianst in the Q7 thread.

Look at the CPU stats for your CPU.  Look at the GPU stats for your GPU.  Consider what "State of the Art" means.  If a company can make a 14 or 16 nm chip but the production costs are orders of magnitude higher; this is not a desirable option unless it is cost effective.  

When Bitcoin Application Specific Integrated Circuits first began to appear there where many chip manufacturers.  This was due to the huge profit margin to be had.  That is long gone; we are in the plateau.  I do think network difficulty will increase with each new generation of miners; but not like it did in 2013 and 2014.  Probably somewhere around 100 PH per generation; with old miner deactivation trailing that increase.

The normal profit model for Bitcoin ASIC has been dependent on the ability to use older tech to make comparatively cheap chips for their mining value.  This was possible because of both a high BTC exchange rate and the disparity between chips a company can cost effectively make and their relative efficiency to those already in existence.  This disparity is what enabled the violation of Moore's law.  The closing of this gap occurred in 2014.  The relative depression of BTC exchange rate in 2015 is why you are only now seeing new 28 nm chips being sold.  I don't believe Bitmain can double efficiency again with a 28 nm chip ( hats off to them if they can ).  Unless the BTC exchange rate goes up significantly I don't believe it is cost effective to produce a 14 or 16 nm asic.  

KnC and Bitfury may roi with their own because they have been mining with them longer; but maybe not.  I think both of them are more interested in becoming the visa and mastercard of the future via new sidechains and XT add ins.  Mining pools will be not only the payment processors of the future; but "processors" in markets that do not currently exist.

I agree the plateau was not do to collusion.  I do not think there is mass collusion between the big three.  However, I do think one of them has been following a very specific hashrate optimization for ~2 years now.  If you have a very large % of the network hashrate, hashing at your maximum possible hashrate continuously is not optimal.  This is because of how the network adjusts the difficulty ( see difficulty explanation ).  The mathematics to optimize complex iterative systems is very counter intuitive.  

I will try to make a simple analogy.  The returns from making better mining equipment is diminishing and approaches zero; thus the graph of making better equipment ( or even more equipment ) vs return ( generated BTC ) is asymptotic. ( it gets closer and closer to zero without ever getting to zero; see https://en.wikipedia.org/wiki/Zeno's_paradoxes )

It is important to understand how the iterative nature of the system makes mining BTC significantly different than say drilling for oil.  It IS possible with enough equipment to drill all the earths reserves of oil in a period of 20 years.  It is NOT possible with infinite equipment to mine all the BTC in a period of 20 years.

Now lets think about how they are the same.  Even though we COULD drill all the earths reserves of oil in a period of 20 years the production AND storage of oil is LIMITED and TIGHTLY CONTROLLED.  Why?  This is how you maximize profit.  Diamonds are similar.  There exists a company which mines almost all diamonds; and holds an enormous horde of them.  They slowly sell them and thus maximize their profit.  

Because no mining company has a monopoly their iteration is modeled after a modified Iterated Prisoner's Dilemma.  It is modified because if they can bankrupt one of the other companies they remove an opponent from the following iterations.

[philipma1957]

The REAL reasons for slower growth rate of late are
(1) the bitcoin price collapse, giving much less incentive to folks to mine and lower profits for those that DO mine and sellers of mining gear
(2) the tech race is slowing a TON as the tech used to mine Bitcoin gets close to the best AVAILABLE tech
(3) the COST of developing new tech gets a LOT higher as you get close to the "bleeding edge" on tech
(4) the death of some miner companies due to the above factors making them unprofitable

 There is ZERO "economic need" to pace growth rate, in fact there is at least one good reason to NOT pace growth (ref market share) for any single company, and there have been way too many cases in history where it's hasn't happened even where it was a GOOD IDEA to do so (look at the history of Oil Production for a well-documented example).

 Do keep in mind that there is a significant random factor in hashrates.

Strange that you now seem to have the same opinions you argued agianst in the Q7 thread.

Look at the CPU stats for your CPU.  Look at the GPU stats for your GPU.  Consider what "State of the Art" means.  If a company can make a 14 or 16 nm chip but the production costs are orders of magnitude higher; this is not a desirable option unless it is cost effective.  

When Bitcoin Application Specific Integrated Circuits first began to appear there where many chip manufacturers.  This was due to the huge profit margin to be had.  That is long gone; we are in the plateau.  I do think network difficulty will increase with each new generation of miners; but not like it did in 2013 and 2014.  Probably somewhere around 100 PH per generation; with old miner deactivation trailing that increase.

The normal profit model for Bitcoin ASIC has been dependent on the ability to use older tech to make comparatively cheap chips for their mining value.  This was possible because of both a high BTC exchange rate and the disparity between chips a company can cost effectively make and their relative efficiency to those already in existence.  This disparity is what enabled the violation of Moore's law.  The closing of this gap occurred in 2014.  The relative depression of BTC exchange rate in 2015 is why you are only now seeing new 28 nm chips being sold.  I don't believe Bitmain can double efficiency again with a 28 nm chip ( hats off to them if they can ).  Unless the BTC exchange rate goes up significantly I don't believe it is cost effective to produce a 14 or 16 nm asic.  

KnC and Bitfury may roi with their own because they have been mining with them longer; but maybe not.  I think both of them are more interested in becoming the visa and mastercard of the future via new sidechains and XT add ins.  Mining pools will be not only the payment processors of the future; but "processors" in markets that do not currently exist.

I agree the plateau was not do to collusion.  I do not think there is mass collusion between the big three.  However, I do think one of them has been following a very specific hashrate optimization for ~2 years now.  If you have a very large % of the network hashrate, hashing at your maximum possible hashrate continuously is not optimal.  This is because of how the network adjusts the difficulty.  The mathematics to optimize complex iterative systems is very counter intuitive.  

I will try to make a simple analogy.  The returns from making better mining equipment is diminishing and approaches zero; thus the graph of making better equipment ( or even more equipment ) vs return ( generated BTC ) is asymptotic. ( it gets closer and closer to zero without ever getting to zero; see https://en.wikipedia.org/wiki/Zeno's_paradoxes )

It is important to understand how the iterative nature of the system makes mining BTC significantly different than say drilling for oil.  It IS possible with enough equipment to drill all the earths reserves of oil in a period of 20 years.  It is NOT possible with infinite equipment to mine all the BTC in a period of 20 years.

Now lets think about how they are the same.  Even though we COULD drill all the earths reserves of oil in a period of 20 years the production AND storage of oil is LIMITED and TIGHTLY CONTROLLED.  Why?  This is how you maximize profit.  Diamonds are similar.  There exists a company which mines almost all diamonds; and holds an enormous horde of them.  They slowly sell them and thus maximize their profit.  

Because no mining company has a monopoly their iteration is modeled after a modified Iterated Prisoner's Dilemma.  It is modified because if they can bankrupt one of the other companies they remove an opponent from the following iterations.

Nice write up.  Pretty much spot on.  Now if a new magic chip is invented like the gpu to asic chip jump
That was from 333 watts a gh to 10 watts a gh we will have big growth.
But .5 watts to .25 watts you get swap and not 100 % swap .
This does give the company a safety margin. It also slows growth.

[QuintLeo]

I would say before they sold the used S5's they likely had the chip design ready for S7's and likely working prototypes.   I thought you had a inside source or saw something we didn't the way you worded you saw something.   

It's really speculating not "seeing" i think that you are doing.   Just as I am doing.

 I'm pretty sure the "used" S5s they were selling had already been replaced, or were in process of getting replaced, by S7 units, and a few of the early ones may have been replaced by S5+ units.

 I can't say for 100% certain though, since I'm not working at Bitmain's farm facility(ies).

Strange that you now seem to have the same opinions you argued agianst in the Q7 thread.

 I do? I've NEVER said anything about anyone deliberately restraining diff increases. What the heck are you talking about?

It IS possible with enough equipment to drill all the earths reserves of oil in a period of 20 years.

 Untill the next oil technology breakthrough. Oil also has major economic limitations on how PROFITABLE it would be to drill it all out that fast, which is why it's never happened and is why we're not likely to "run out" of oil for a good century or more (though we might see more short-term shortfalls as we have done in the past).

 There is some similarity of oil and bitcoin though, in that they are both limited in available product TO "mine" - the biggest difference is the artificial limits on how fast Bitcoin CAN be mined, but it still makes sense for a big miner (be that a manufacturer or not) to push to get as many as it can now BEFORE OTHERS BEAT IT TO THEM, and makes ZERO sense for a miner to try to "hold back the hashrate" - there are too many folks that WON'T collude to make that a viable strategy, unlike oil production where the number of significant producers is strictly limted by the HUGE PRICE OF ENTRY and the limitation of most major oil producing nations having nationalised their oil and access to it (The US and UK are the only significant exceptions to that last, and even the UK has been semi-nationalised for a lot of the time it's oil industry has existed).

 P.S. - who do you consider the "big three" - there's more than 3 makers still around, though a few don't sell to the public any more.

 D'Beers (sp?) is a very different situation from oil. One company in one country controlling the majority of the "reserves" of Diamonds is in no way comparable to the number of producers and countries in the Oil industry - even the Saudis and their "largest proven reserves of any country in the world" are nowhere near having the kind of control D'Beers has over diamonds (witness the current "oil price war", which certainly is NOT maximising profits for anyone).

Now if a new magic chip is invented like the gpu to asic chip jump

 Closest we'll see to that sort of thing would be a "full custom" chip in 14/16nm - at this point the cryptocoin ASIC state of the art is too close to the leading edge of semiconductor technology in general for more than 1 more major efficiency jump in the next few years. The big jumps in ASIC have been about playing "catch-up to the state of the art", and there's only one more such jump available at this point and for the next few years.

[lucky1974]

I would say before they sold the used S5's they likely had the chip design ready for S7's and likely working prototypes.   I thought you had a inside source or saw something we didn't the way you worded you saw something.    

It's really speculating not "seeing" i think that you are doing.   Just as I am doing.

 I'm pretty sure the "used" S5s they were selling had already been replaced, or were in process of getting replaced, by S7 units, and a few of the early ones may have been replaced by S5+ units.

 I can't say for 100% certain though, since I'm not working at Bitmain's farm facility(ies).

Strange that you now seem to have the same opinions you argued agianst in the Q7 thread.

 I do? I've NEVER said anything about anyone deliberately restraining diff increases. What the heck are you talking about?

It IS possible with enough equipment to drill all the earths reserves of oil in a period of 20 years.

 Untill the next oil technology breakthrough. Oil also has major economic limitations on how PROFITABLE it would be to drill it all out that fast, which is why it's never happened and is why we're not likely to "run out" of oil for a good century or more (though we might see more short-term shortfalls as we have done in the past).

 There is some similarity of oil and bitcoin though, in that they are both limited in available product TO "mine" - the biggest difference is the artificial limits on how fast Bitcoin CAN be mined, but it still makes sense for a big miner (be that a manufacturer or not) to push to get as many as it can now BEFORE OTHERS BEAT IT TO THEM, and makes ZERO sense for a miner to try to "hold back the hashrate" - there are too many folks that WON'T collude to make that a viable strategy, unlike oil production where the number of significant producers is strictly limted by the HUGE PRICE OF ENTRY and the limitation of most major oil producing nations having nationalised their oil and access to it (The US and UK are the only significant exceptions to that last, and even the UK has been semi-nationalised for a lot of the time it's oil industry has existed).

 P.S. - who do you consider the "big three" - there's more than 3 makers still around, though a few don't sell to the public any more.

 D'Beers (sp?) is a very different situation from oil. One company in one country controlling the majority of the "reserves" of Diamonds is in no way comparable to the number of producers and countries in the Oil industry - even the Saudis and their "largest proven reserves of any country in the world" are nowhere near having the kind of control D'Beers has over diamonds (witness the current "oil price war", which certainly is NOT maximising profits for anyone).

Now if a new magic chip is invented like the gpu to asic chip jump

 Closest we'll see to that sort of thing would be a "full custom" chip in 14/16nm - at this point the cryptocoin ASIC state of the art is too close to the leading edge of semiconductor technology in general for more than 1 more major efficiency jump in the next few years. The big jumps in ASIC have been about playing "catch-up to the state of the art", and there's only one more such jump available at this point and for the next few years.

all this philosophy especially from quintleo lol lol lol