Projected Minimum Cost per BTC over the next year

[BurtW]

This thread is just a quick attempt to project the minimum cost per BTC over the next year assuming the growth in the hash rate next year matches the growth in the hash rate over the last year.

tl;dr:  either the growth in the hash rate must slow down, the power consumption must go down, or the price of BTC must go up, a lot.

Using the following conversion factors, constants and assumptions:

Code:

GH/s per Diff   0.007158388055
Blocks/Period   2016          
BTC/Period      50400         
Watts per GH/s  1 (assumed constant for the next year - it may go down)             
USD/kWh         $0.10         

In other words assuming everyone in the network pays $0.10 per kWh and everyone has miners that burn 1 W per GH/s (1 J/GH) then we can calculate the average production cost for each BTC over the last year as follows:

Code:

                                         Hash Rate   Power     Energy      Cost    Cost
     Date     Difficulty   Delta Days        GH/s      kW        kWh  $/Period   $/BTC
---------  -------------  ------ ----  ----------  ------  ---------  --------  ------
14-Mar-13      4,847,647                   34,701      35                             
24-Mar-13      6,695,826  38.13%   10      47,931      48     11,504    $1,150   $0.02
05-Apr-13      7,673,000  14.59%   12      54,926      55     15,819    $1,582   $0.03
17-Apr-13      8,974,296  16.96%   12      64,241      64     18,502    $1,850   $0.04
30-Apr-13     10,076,293  12.28%   13      72,130      72     22,505    $2,250   $0.04
12-May-13     11,187,257  11.03%   12      80,083      80     23,064    $2,306   $0.05
25-May-13     12,153,412   8.64%   13      86,999      87     27,144    $2,714   $0.05
05-Jun-13     15,605,633  28.41%   11     111,711     112     29,492    $2,949   $0.06
16-Jun-13     19,339,258  23.92%   11     138,438     138     36,548    $3,655   $0.07
29-Jun-13     21,335,329  10.32%   13     152,727     153     47,651    $4,765   $0.09
11-Jul-13     26,162,876  22.63%   12     187,284     187     53,938    $5,394   $0.11
22-Jul-13     31,256,961  19.47%   11     223,749     224     59,070    $5,907   $0.12
03-Aug-13     37,392,766  19.63%   12     267,672     268     77,090    $7,709   $0.15
13-Aug-13     50,810,339  35.88%   10     363,720     364     87,293    $8,729   $0.17
24-Aug-13     65,750,060  29.40%   11     470,664     471    124,255   $12,426   $0.25
04-Sep-13     86,933,018  32.22%   11     622,300     622    164,287   $16,429   $0.33
14-Sep-13    112,628,549  29.56%   10     806,239     806    193,497   $19,350   $0.38
25-Sep-13    148,819,200  32.13%   11   1,065,306   1,065    281,241   $28,124   $0.56
06-Oct-13    189,281,249  27.19%   11   1,354,949   1,355    357,706   $35,771   $0.71
16-Oct-13    267,731,249  41.45%   10   1,916,524   1,917    459,966   $45,997   $0.91
26-Oct-13    390,928,788  46.02%   10   2,798,420   2,798    671,621   $67,162   $1.33
05-Nov-13    510,929,738  30.70%   10   3,657,433   3,657    877,784   $87,778   $1.74
17-Nov-13    609,482,680  19.29%   12   4,362,914   4,363  1,256,519  $125,652   $2.49
29-Nov-13    707,408,283  16.07%   12   5,063,903   5,064  1,458,404  $145,840   $2.89
10-Dec-13    908,350,862  28.41%   11   6,502,328   6,502  1,716,615  $171,661   $3.41
21-Dec-13  1,180,923,195  30.01%   11   8,453,506   8,454  2,231,726  $223,173   $4.43
02-Jan-14  1,418,481,395  20.12%   12  10,154,040  10,154  2,924,364  $292,436   $5.80
13-Jan-14  1,789,546,951  26.16%   11  12,810,272  12,810  3,381,912  $338,191   $6.71
24-Jan-14  2,193,847,870  22.59%   11  15,704,414  15,704  4,145,965  $414,597   $8.23
05-Feb-14  2,621,404,453  19.49%   12  18,765,030  18,765  5,404,329  $540,433  $10.72
17-Feb-14  3,129,573,175  19.39%   12  22,402,699  22,403  6,451,977  $645,198  $12.80
28-Feb-14  3,815,723,799  21.92%   11  27,314,432  27,314  7,211,010  $721,101  $14.31
13-Mar-14  4,250,217,920  11.39%   13  30,424,709  30,425  9,492,509  $949,251  $18.83

Continued in next post...

[1714865182]

1714865182

[1714865182]

1714865182

[1714865182]

1714865182

[BurtW]

Over the last year the average increase in the difficulty and hash rate was 23.92% per adjustment period and the average length of each adjustment period was 11.38 days.

Assuming the network growth rate over the next year is the same as it was this last year we get:

Code:

                                Hash Rate   Power     Energy          Cost         Cost
     Date         Difficulty        TH/s      MW        MWh      $/Period        $/BTC
---------  -----------------  ----------  ------  ---------  ------------   ----------
24-Mar-14      5,266,766,872      37,702      38     10,293    $1,029,253       $20.42
04-Apr-14      6,526,449,656      46,719      47     12,754    $1,275,425       $25.31
16-Apr-14      8,087,417,983      57,893      58     15,805    $1,580,476       $31.36
27-Apr-14     10,021,732,041      71,739      72     19,585    $1,958,487       $38.86
08-May-14     12,418,687,065      88,898      89     24,269    $2,426,909       $48.15
20-May-14     15,388,935,546     110,160     110     30,074    $3,007,367       $59.67
31-May-14     19,069,595,359     136,508     137     37,267    $3,726,656       $73.94
12-Jun-14     23,630,579,651     169,157     169     46,180    $4,617,982       $91.63
23-Jun-14     29,282,440,667     209,615     210     57,225    $5,722,492      $113.54
04-Jul-14     36,286,089,639     259,750     260     70,912    $7,091,173      $140.70
16-Jul-14     44,964,841,429     321,876     322     87,872    $8,787,209      $174.35
27-Jul-14     55,719,339,970     398,861     399    108,889   $10,888,896      $216.05
07-Aug-14     69,046,053,494     494,258     494    134,933   $13,493,256      $267.72
19-Aug-14     85,560,193,384     612,473     612    167,205   $16,720,515      $331.76
30-Aug-14    106,024,114,653     758,962     759    207,197   $20,719,656      $411.10
11-Sep-14    131,382,509,124     940,487     940    256,753   $25,675,295      $509.43
22-Sep-14    162,806,015,974   1,165,429   1,165    318,162   $31,816,202      $631.27
03-Oct-14    201,745,262,850   1,444,171   1,444    394,259   $39,425,865      $782.26
15-Oct-14    249,997,832,322   1,789,581   1,790    488,556   $48,855,575      $969.36
26-Oct-14    309,791,245,073   2,217,606   2,218    605,406   $60,540,642    $1,201.20
06-Nov-14    383,885,790,658   2,748,003   2,748    750,205   $75,020,494    $1,488.50
18-Nov-14    475,701,952,889   3,405,259   3,405    929,636   $92,963,576    $1,844.52
29-Nov-14    589,478,312,272   4,219,715   4,220  1,151,982  $115,198,206    $2,285.68
11-Dec-14    730,467,214,878   5,228,968   5,229  1,427,508  $142,750,821    $2,832.36
22-Dec-14    905,177,240,458   6,479,610   6,480  1,768,934  $176,893,352    $3,509.79
02-Jan-15  1,121,673,663,040   8,029,375   8,029  2,192,019  $219,201,947    $4,349.24
14-Jan-15  1,389,950,774,416   9,949,807   9,950  2,716,297  $271,629,732    $5,389.48
25-Jan-15  1,722,393,258,358  12,329,559  12,330  3,365,970  $336,596,970    $6,678.51
05-Feb-15  2,134,347,914,361  15,278,491  15,278  4,171,028  $417,102,794    $8,275.85
17-Feb-15  2,644,832,123,808  18,932,735  18,933  5,168,637  $516,863,657   $10,255.23
28-Feb-15  3,277,411,764,061  23,460,985  23,461  6,404,849  $640,484,897   $12,708.03
12-Mar-15  4,061,289,098,283  29,072,283  29,072  7,936,733  $793,673,336   $15,747.49

[chriswilmer]

Cool!

OK, so, assuming efficiency doubles... we can halve the cost of bitcoins produced in our heads pretty easily...

Just to be crystal clear, you are assuming 25 btc per block with no transaction fees right?

[chriswilmer]

Also, it's interesting to keep in mind that these are variable costs... and that the capital costs are quite significant.

It would be interesting to do a calculation that lumped amortized capital costs in together with the variable costs to get a better sense of much bitcoins costs to produce in practical terms. (clearly this would be much more subjective... but as long as the parameters were easy for people to play with...)

[BurtW]

Cool!

OK, so, assuming efficiency doubles... we can halve the cost of bitcoins produced in our heads pretty easily...

Just to be crystal clear, you are assuming 25 btc per block with no transaction fees right?

But, these numbers already assume the current best case efficiency and assume that the entire network is only burning 1J/GH.  This is not the case.  The network currently burns a lot more than that because not everyone has the most efficient, up to date, latest and greatest miners.  See Greg's thread here: https://bitcointalk.org/index.php?topic=281279.0 which shows that the most efficient miners on the market are at about 0.8 J/GH plus overhead which is why I used 1 J/GH as a best case estimate.

Yes, in the assumptions I show 50,400 BTC per adjustment period.  Each adjustment period is exacly 2,016 blocks so these numbers assume 50,400 / 2,016 = 25 BTC per block.  I do not take fees into consideration because they are small compared to the block reward and I am trying to calculate the USD and energy cost of producing new BTC.

[BurtW]

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

Shows the 2008 worldwide power generation at 2,311.4 GW (20,261 TWh of energy).  

If the current growth rate continues then one year from now the entire Bitcoin network will consume a minimum of 29,072 MW or 29.072 / 2,311.4 = 1.26% of all electrical power generated on the planet and it will cost a minimum of $15,747 to produce each new BTC.

Code:

                                Hash Rate   Power     Energy          Cost         Cost
     Date         Difficulty        TH/s      MW        MWh      $/Period        $/BTC
---------  -----------------  ----------  ------  ---------  ------------   ----------
12-Mar-15  4,061,289,098,283  29,072,283  29,072  7,936,733  $793,673,336   $15,747.49

So...

I am betting on a leveling off of the growth rate this year.

[chriswilmer]

I follow ASICMINER pretty closely and they are coming out with 0.35 J/GH miners imminently. *shrugs* That's not confirmed until someone has a working miner in front of them... but I wouldn't confidently exclude the possibility. I think we will probably see 0.1 J/GH sometime in 2015.

[BurtW]

I follow ASICMINER pretty closely and they are coming out with 0.35 J/GH miners imminently. *shrugs* That's not confirmed until someone has a working miner in front of them... but I wouldn't confidently exclude the possibility. I think we will probably see 0.1 J/GH sometime in 2015.

Sure, if either of those things happen then the numbers will adjust accordingly.  However, I have to wonder why the 0.35 J/GH specification has not been included in Greg's thread.  Perhaps he has stopped updating it?  If you have a link to an actual spec and you can post it in that thread then maybe we can get him to update the OP in that thread.

[pdawg]

This is why the next 12-24 months will be very, very interesting for BTC.  I do think we are already seeing a leveling off to a certain degree of hash power, probably somewhat related to the recent price drop which affects for-profit farms (i.e., powering down if price is below a certain threshold).  But the very large farms are getting equipment and electricity for very cheap so that scenario does not apply to them unless there is a severe price crash (which I don't see happening due to the broad adoption phase we are currently in).  I can't help but believe that BTC value will sky rocket due to the raw cost to mine and the lack of new supply being introduced into the marketplace.

[chriswilmer]

I follow ASICMINER pretty closely and they are coming out with 0.35 J/GH miners imminently. *shrugs* That's not confirmed until someone has a working miner in front of them... but I wouldn't confidently exclude the possibility. I think we will probably see 0.1 J/GH sometime in 2015.

Sure, if either of those things happen then the numbers will adjust accordingly.  However, I have to wonder why the 0.35 J/GH specification has not been included in Greg's thread.  Perhaps he has stopped updating it?  If you have a link to an actual spec and you can post it in that thread then maybe we can get him to update the OP in that thread.

Here are the chip specs:

https://bitcointalk.org/index.php?topic=438359.0

[BorisAlt]

Burt, could you re-compute please for 0.35w/h case?

[chriswilmer]

Burt, could you re-compute please for 0.35w/h case?

I think you can just multiply all of the numbers on the right by 0.35

[mwizard]

I find the initial assumption of this thread that there would be a continuing 23.92% difficulty increase completely unrealistic.

I have 2 reasons.

1) There are no obvious technical reasons to expect large hash rate improvements in the next year.

This time last year the first ASIC systems where shipping and it was obvious to everyone that  difficulty would zoom up.   People were still using GPUs with a hash rate of one megahash or so. The entire world hashing rate in March last year was just 35 terrahash.  

The mining hardware world is now much more stable with no major technological advances expected.   The only expected change are a movement to 20nm with some Joule/Gigahash improvements (as mentioned earlier in this thread).

2) The size of the hashing network is now so large it is difficulty to grow by such large amounts.  Right now a 23.92% increase would require adding a petahash (1,000 terrahash) each day and even more in the future.  It is just not happening.  The sheer size of the current hashing network makes further growth more difficult.

It would be a much more realistic to assume an average 15% increase till mid 2014 giving a hashing network size of about 120 petahash by June 2014,  10% average till the end of the year giving 300 petahash, and 5% during 2015 giving 500 petahash at some time during 2015.  This still means enormous network growth with 20+ petahash being added per fortnight during 2015.

[seriouscoin]

Wow for a hero member, with activity over 1000 make a complete joke trial at projecting cost of BTC......unreal....


I cant stop laughing.... I will let someone explain to you a thing or two about production cost....

Hint: just taking assumed electrical cost of the mining network is NOT how you calculate the production cost of BTC.
 

[BorisAlt]

Wow for a hero member, with activity over 1000 make a complete joke trial at projecting cost of BTC......unreal....


I cant stop laughing.... I will let someone explain to you a thing or two about production cost....

Hint: just taking assumed electrical cost of the mining network is NOT how you calculate the production cost of BTC.
 

How do you calculate? Why will you let someone else to explain if you know?

[BurtW]

I am glad you got a good laugh at me and gmaxwell since his thread does exactly the same thing from a different angle.  Obviously I was not trying to calculate the actual production cost.  You seem to have missed the entire point of the thread.

Since you missed it others might also so I have changed the title and OP.

[Bicknellski]

Any planned updates to the thread BurtW?

[samsonn25]

What I do know is the difficulty is 1000x what it was last spring and the btc price is 5x higher.

[BurtW]

There really is nothing new to report.  The difficulty continues to climb at about the same rate and the price continues to go down.  This wil not last.  One of these things MUST give.  Either the price needs to go up or the difficulty must stop going up.  Eventually.

I will redo the calculations in a month or two.

[cloverme]

There really is nothing new to report.  The difficulty continues to climb at about the same rate and the price continues to go down.  This wil not last.  One of these things MUST give.  Either the price needs to go up or the difficulty must stop going up.  Eventually.

I will redo the calculations in a month or two.

The other issue is that all bitcoins are equal, btc mined at a lower energy cost are worth as much as those generated at a higher energy cost relative to difficulty. Eventually, like you correctly predict, the market will not be able to sustain the two factors going in the opposite direction. I think the balance will occur when the supply of BTC goes down as difficulty rises or BTC will crash in price and large hashing farms stop mining because it won't be profitable. As a result, the hashrate would drop and difficulty with it.  I think we are going to see mining have some sort of a correction this summer unless price goes up by 40%.