Total Hashrate Forecast Q1, Q2 2014 (Community work)

[ecliptic]

Any clue where this 5PH in ~10 days is from?

[darcimer]

What are the options?
the Antminer family of products
Cointerra testing
Private hashing

I just hope it's something we already factored in..

[Maddin]

I did also a forecast, but only based on history:


Development of difficulty since 15.09.13
Dates   days to solve 2016   difficulty    increased in %
15.09.13- 26.09.13   11   112628549   --
26.09.13-7.10.13   11   148819200   32,13
7.10.13-17.10.13   10   189281249   27,19
17.10.13-26.10.13   9   267731249   41,48
26.10.13-06.11.13   11   390928788   46,02
6.11.13-18.11.13   12   510929738   30,07
18.11.13-30.11.13   12   609482680   19,29
30.11.13-11.12.13   11   707408283   16,07
11.12.13-22.12.13   11   908350862   28,41
22.13.13-03.01.14   12   1180923195   30,01
03.01.14-14.01.14   11   1418481395   20,12
14.01.14- ??              1789546951      26,15

= average time to solve 2016 Blocks: 121/11 = 11
= average difficulty increase: 316,94 / 11 = 28,81 %

Forecast
dates    days for solving 2016   difficulty    increased in %
14.01.14-25.01.14    11   1789546951   26,15
25.01.14-05.02.14    11   2305115427   28,81
                            11   2969219182   28,81
                            11   3824651228   28,81
                            11   4926533247   28,81
                            11   6345867476   28,81
                            11   8174111896   28,81
                            11   10529073533   28,81
                            11   13562499618   28,81
                                  11   17469855758   28,81
                            11   22502921202   28,81
                            11   28986012801   28,81
                            11   37336883088   28,81
17.06.14-28.06.14      11   48093639106   28,81

I was to lazy to put in all the dates :-D

[steamon]

Any clue where this 5PH in ~10 days is from?

Problem is that most people and also richer people saw the 1k btc price and think they can get even more richer. If a few off those start self producing or private orders yea then we can see even crazyer increasements. Lets hope everything is calculated here because thats were me own investment off the long term is based on.

[novello]

I did also a forecast, but only based on history:


Development of difficulty since 15.09.13
Dates   days to solve 2016   difficulty    increased in %
15.09.13- 26.09.13   11   112628549   --
26.09.13-7.10.13   11   148819200   32,13
7.10.13-17.10.13   10   189281249   27,19
17.10.13-26.10.13   9   267731249   41,48
26.10.13-06.11.13   11   390928788   46,02
6.11.13-18.11.13   12   510929738   30,07
18.11.13-30.11.13   12   609482680   19,29
30.11.13-11.12.13   11   707408283   16,07
11.12.13-22.12.13   11   908350862   28,41
22.13.13-03.01.14   12   1180923195   30,01
03.01.14-14.01.14   11   1418481395   20,12
14.01.14- ??              1789546951      26,15

= average time to solve 2016 Blocks: 121/11 = 11
= average difficulty increase: 316,94 / 11 = 28,81 %

Forecast
dates    days for solving 2016   difficulty    increased in %
14.01.14-25.01.14    11   1789546951   26,15
25.01.14-05.02.14    11   2305115427   28,81
                            11   2969219182   28,81
                            11   3824651228   28,81
                            11   4926533247   28,81
                            11   6345867476   28,81
                            11   8174111896   28,81
                            11   10529073533   28,81
                            11   13562499618   28,81
                                  11   17469855758   28,81
                            11   22502921202   28,81
                            11   28986012801   28,81
                            11   37336883088   28,81
17.06.14-28.06.14      11   48093639106   28,81

I was to lazy to put in all the dates :-D

We've spent a lot of time on a Behavioral Model to try to make some sense of what's going on. It pretty soon became clear that there is tipping point where miners will stop buying new kit, because they simply will never make a profit. The difficulty can't keep going up exponentially as a result (unless the Bitcoin value follows the same trend).Our prediction shows difficulty in mid-June at around 18,000,000,000,  25,000,000,000 by end 2014. At that time a TH will earn about $400 a month and eat a good portion of that in electricity costs. As a result, it's worth less than $1.3 per Gigahash/sec.

[Entropy-uc]

I did also a forecast, but only based on history:


Development of difficulty since 15.09.13
Dates   days to solve 2016   difficulty    increased in %
15.09.13- 26.09.13   11   112628549   --
26.09.13-7.10.13   11   148819200   32,13
7.10.13-17.10.13   10   189281249   27,19
17.10.13-26.10.13   9   267731249   41,48
26.10.13-06.11.13   11   390928788   46,02
6.11.13-18.11.13   12   510929738   30,07
18.11.13-30.11.13   12   609482680   19,29
30.11.13-11.12.13   11   707408283   16,07
11.12.13-22.12.13   11   908350862   28,41
22.13.13-03.01.14   12   1180923195   30,01
03.01.14-14.01.14   11   1418481395   20,12
14.01.14- ??              1789546951      26,15

= average time to solve 2016 Blocks: 121/11 = 11
= average difficulty increase: 316,94 / 11 = 28,81 %

Forecast
dates    days for solving 2016   difficulty    increased in %
14.01.14-25.01.14    11   1789546951   26,15
25.01.14-05.02.14    11   2305115427   28,81
                            11   2969219182   28,81
                            11   3824651228   28,81
                            11   4926533247   28,81
                            11   6345867476   28,81
                            11   8174111896   28,81
                            11   10529073533   28,81
                            11   13562499618   28,81
                                  11   17469855758   28,81
                            11   22502921202   28,81
                            11   28986012801   28,81
                            11   37336883088   28,81
17.06.14-28.06.14      11   48093639106   28,81

I was to lazy to put in all the dates :-D

We've spent a lot of time on a Behavioral Model to try to make some sense of what's going on. It pretty soon became clear that there is tipping point where miners will stop buying new kit, because they simply will never make a profit. The difficulty can't keep going up exponentially as a result (unless the Bitcoin value follows the same trend).Our prediction shows difficulty in mid-June at around 18,000,000,000,  25,000,000,000 by end 2014. At that time a TH will earn about $400 a month and eat a good portion of that in electricity costs. As a result, it's worth less than $1.3 per Gigahash/sec.

Due to the preorder cycle, it will take 3-4 months for the feedback of difficulty to impact buying behavior. So difficulty continues ramping up way beyond your expectation, and prices in the secondary markers collapse.  The overshoot might be moderated by a major supplier going bankrupt and not delivering but otherwise expect your year end prediction to be reached in July.

[omehenk]

Any clue where this 5PH in ~10 days is from?

Problem is that most people and also richer people saw the 1k btc price and think they can get even more richer. If a few off those start self producing or private orders yea then we can see even crazyer increasements. Lets hope everything is calculated here because thats were me own investment off the long term is based on.

I know people with a lot off Money, they are indeed Producing there own hardware.
Only then its still highly profitable .

And in countries with cheap electricity . 

For the coming, 5 times atleast   30 % + up.


H

[novello]

Due to the preorder cycle, it will take 3-4 months for the feedback of difficulty to impact buying behavior. So difficulty continues ramping up way beyond your expectation, and prices in the secondary markers collapse.  The overshoot might be moderated by a major supplier going bankrupt and not delivering but otherwise expect your year end prediction to be reached in July.

You're making assumptions about assumptions hereon buying behavior based on the huge price jump in Q4 last year. You can't automatically factor in that happening again, so as soon as the difficulty hits about 15000 x 10e6 the current (yet to be delivered) $3/GH rigs will never be profitable and it's not going to take a genius to see the trend,

The companies that have their own asics are in a slightly different boat but even they have to look at RI and they might not be quite so keen to load in capacity which might not make their ROI target.

Time will tell what approach is right, neither of us have a crystal ball.

[Entropy-uc]

Due to the preorder cycle, it will take 3-4 months for the feedback of difficulty to impact buying behavior. So difficulty continues ramping up way beyond your expectation, and prices in the secondary markers collapse.  The overshoot might be moderated by a major supplier going bankrupt and not delivering but otherwise expect your year end prediction to be reached in July.

You're making assumptions about assumptions hereon buying behavior based on the huge price jump in Q4 last year. You can't automatically factor in that happening again, so as soon as the difficulty hits about 15000 x 10e6 the current (yet to be delivered) $3/GH rigs will never be profitable and it's not going to take a genius to see the trend,

The companies that have their own asics are in a slightly different boat but even they have to look at RI and they might not be quite so keen to load in capacity which might not make their ROI target.

Time will tell what approach is right, neither of us have a crystal ball.

Time may tell, but I can already tell who is talking his book.

Hello folks, greetings from Novello Technologies Ltd in the UK.

We'll shortly be launching a funding campaign to finance the development of a family of low cost, high performance mining rigs.

How does $1 per Gigahash/second (or less) sound to you? But before you groan "not another mning asic startup" wait until you see our plan, it's not what you might expect.

Gordon

[novello]

Due to the preorder cycle, it will take 3-4 months for the feedback of difficulty to impact buying behavior. So difficulty continues ramping up way beyond your expectation, and prices in the secondary markers collapse.  The overshoot might be moderated by a major supplier going bankrupt and not delivering but otherwise expect your year end prediction to be reached in July.

You're making assumptions about assumptions hereon buying behavior based on the huge price jump in Q4 last year. You can't automatically factor in that happening again, so as soon as the difficulty hits about 15000 x 10e6 the current (yet to be delivered) $3/GH rigs will never be profitable and it's not going to take a genius to see the trend,

The companies that have their own asics are in a slightly different boat but even they have to look at RI and they might not be quite so keen to load in capacity which might not make their ROI target.

Time will tell what approach is right, neither of us have a crystal ball.

Time may tell, but I can already tell who is talking his book.

Hello folks, greetings from Novello Technologies Ltd in the UK.

We'll shortly be launching a funding campaign to finance the development of a family of low cost, high performance mining rigs.

How does $1 per Gigahash/second (or less) sound to you? But before you groan "not another mning asic startup" wait until you see our plan, it's not what you might expect.

Gordon

Yes, it's not exactly a secret what we plan to do, but we plan to do it right. We have indeed developed a behavioral model to try to predict what the market for rigs might be, but since we won't be delivering anything until August it's no advantage to suggest that some of the estimates other have submitted might be a bit too high. Scare tactics, if you will. Why would anyone want to do  such a thing?

If we're suggesting that the difficulty might not be so high, it's going to benefit our competitors (assuming we actually get to the competing stage and that's not a foregone conclusion) who will get more people ordering if they think they might have a chance of profitability on their purchase.

So I'm a bit unclear what book you think we might be talking/selling. One that tells the truth, perhaps?

[Maddin]

I think theres a issue with your predictions: You assume that the miners act rational. I think its more like: "what, I can produce money? sounds good! lets do that!" If people would act rational, the grow of the hashrate would have stopped long ago. So I think that the hashrate will keep growing exponentially and therefore pushing the price of bitcoins up.

[ssbn506]

I think theres a issue with your predictions: You assume that the miners act rational. I think its more like: "what, I can produce money? sounds good! lets do that!" If people would act rational, the grow of the hashrate would have stopped long ago. So I think that the hashrate will keep growing exponentially and therefore pushing the price of bitcoins up.

Another thing to consider is the endless pool of new people getting into the game. I wish I knew what percentage of products the miner manufactures sell are repeat customers. Most business rely on something like 80% repeat business. If this is the case for mining hardware companies they will quickly go out of business.(quickly like in 2 to 5 years) But I think the reality is more like 20% repeat and 80% new with a huge lineup. The lineup of people who cant do math and think this is get rich quick is huge for now. But bitcoin is still a tiny market and something like 1% of users holding 80% of BTC. So in the long run it may be good. This is just the growing pains of a new market.

[NicholasB54]

Here's another Cointerra data source
http://cointerra.com/genisis-block-cointerra-hashfast-sell-17m-bitcoin-miners-open-second-batch-sales/

Great topic, I see Bitmain mentioned (Antminers) but not included in the forecast - are they low volume

[crazydownloaded]

I tried to summarize all those data into a spreadsheet: https://docs.google.com/spreadsheet/ccc?key=0AqGoWjATRQK6dE1yTHpydkIwajhhUGF4ZGxFV2FDSVE&usp=drive_web#gid=1
This also outputs difficulties estimates on the first tab, which you can then enter on http://coinplorer.com/Hardware/Simulate (select "manual difficulty input") to estimate revenues, etc.

Let me know if you disagree with any data, the spreadsheet is pretty easy to change ^^

[BuildTheFuture]

Nice spreadsheet, it would be interesting just for fun to extend it into the 2nd half of the year, perhaps just starting with the idea that each company will try to gradually increase their sold hash rate each month.

Also in recent new CoinTerra announced their first shipments will not reach 2 TH/s per unit. Looks like each of their units may average around 1.65 TH/s until they can fix their power usage at some future revision.

[crazydownloaded]

Nice spreadsheet, it would be interesting just for fun to extend it into the 2nd half of the year, perhaps just starting with the idea that each company will try to gradually increase their sold hash rate each month.

Sure, do you have any shipping estimate for Q3/Q4? Nothing have been announced yet to my knowledge.

Also in recent new CoinTerra announced their first shipments will not reach 2 TH/s per unit. Looks like each of their units may average around 1.65 TH/s until they can fix their power usage at some future revision.

Yep, corrected. 600*1.65=990, rounded to 1000.

[de_ixie]

Thank you everybody for your input

I will do an update with refinded numbers after the next difficulty jump. We will use the new diff. level as the new starting point and we will also include all new information provided.

Cheers

[Sonny]

I will do an update with refinded numbers after the next difficulty jump. We will use the new diff. level as the new starting point and we will also include all new information provided.

Cheers

Thanks a lot for that

[NicholasB54]

I tried to summarize all those data into a spreadsheet: https://docs.google.com/spreadsheet/ccc?key=0AqGoWjATRQK6dE1yTHpydkIwajhhUGF4ZGxFV2FDSVE&usp=drive_web#gid=1
.....
Let me know if you disagree with any data, the spreadsheet is pretty easy to change ^^

Great value in that spreadsheet (& this whole topic)

March looking like the perfect storm for anyone looking to earn a payback on existing hardware - more hashing increase than all other months combined
I know the objective is a prediction for end of June, but since you've done that much work already - could/should it be smoothed out more ?

  5,800 = Feb Total
48,600 = March Total
14,750 = April Total
  5,250 = May Total
  1,500 = June Total
 
Is the front loading accidental or based on real hard data ?
(insofar as any ASIC Miner companies predictions are real/hard data)

EDIT : I'll see if I can do some smoothing out myself - since the next couple of months concerns me just as much as mid-year

[crazydownloaded]

This thread is very useful to estimate quantities shipped, but doesn't tell anything on the - as important - when, so I tried to estimate.
I didn't want to artificially "even" the numbers. Simply let me know if you think that some shipping dates might be wrong.