Mining Estimation Sheet

[Benson Samuel]

I have been working on https://docs.google.com/spreadsheet/ccc?key=0Amirb0HzskGTdEdfM3pzTWdOLUFjZFlYRWJaOVk5YWc&usp=sharing to take a look at some rough numbers in terms of mining estimates for 52 difficulty jumps. Roughly 6 months worth of opinions.

Need some feedback on how to improve this model. This is based on a fixed % increase in difficulty.

Orange fields need to be updated. Everything else should calculate.

[laserwolf]

ah but estimating future difficulty is just like trying to estimate the future price.
if you look at the chart of difficultly and the chart of price, you can see they're very correlated.

but there may be a good model which estimates mining profit in $ rather than BTC. since difficulty and price are correlated, it may be possible to cancel their effects.

[Benson Samuel]

ah but estimating future difficulty is just like trying to estimate the future price.
if you look at the chart of difficultly and the chart of price, you can see they're very correlated.

but there may be a good model which estimates mining profit in $ rather than BTC. since difficulty and price are correlated, it may be possible to cancel their effects.

No chance of getting anything accurate. Just trying to get a sheet that shows me some numbers without much trouble.

No stable historical data to map individual jumps as well.

[rezurect]

Looks good
Offtopic: So a 10Gh/s 50btc ASICminer pays itself in about 4 months.
(In a fixed diff increase model)

[Dashing]

Thanks Benson looks good

Off-Topic (Somewhat related),
I don't get it why people buy the 10 Ghash/sec miner from ASICminer (?) It's around 50 BTC each and well not so efficient either (130 watts?), Hopefully Avalon ships their chips soon and a few people can start to sell their own made ASIC miners with those chips

[Benson Samuel]

Thanks Benson looks good

Off-Topic (Somewhat related),
I don't get it why people buy the 10 Ghash/sec miner from ASICminer (?) It's around 50 BTC each and well not so efficient either (130 watts?), Hopefully Avalon ships their chips soon and a few people can start to sell their own made ASIC miners with those chips

Timing Dashing. Cannot get mass Hashrates from anything else at this point in time.

[Dashing]

Thanks Benson looks good

Off-Topic (Somewhat related),
I don't get it why people buy the 10 Ghash/sec miner from ASICminer (?) It's around 50 BTC each and well not so efficient either (130 watts?), Hopefully Avalon ships their chips soon and a few people can start to sell their own made ASIC miners with those chips

Timing Dashing. Cannot get mass Hashrates from anything else at this point in time.

 Yes, that's pretty true, it's the best hash rate producing device available at the moment

but the odd thing I've noticed is the initial ROI time for the mining blade from asicminer is the same as the time needed for the ROI for a small Lancelot FPGA, around 4 months. Which, is kinda bad for an ASIC, but yeah after the initial 4 months you'd profit pretty well I guess

[rezurect]

Thanks Benson looks good

Off-Topic (Somewhat related),
I don't get it why people buy the 10 Ghash/sec miner from ASICminer (?) It's around 50 BTC each and well not so efficient either (130 watts?), Hopefully Avalon ships their chips soon and a few people can start to sell their own made ASIC miners with those chips

Timing Dashing. Cannot get mass Hashrates from anything else at this point in time.

 Yes, that's pretty true, it's the best hash rate producing device available at the moment

but the odd thing I've noticed is the initial ROI time for the mining blade from asicminer is the same as the time needed for the ROI for a small Lancelot FPGA, around 4 months. Which, is kinda bad for an ASIC, but yeah after the initial 4 months you'd profit pretty well I guess

But it isnt scalable.
A Lancelot costs, what, around 3.5btc including shipping.  
400 megahashes @30watt . It takes 4 months  to make 3.5 btc. For the next 3.5 btc more.
Considering a 3.5 btc investment its ok.
If you try to scale it to 10Gh/s,
Chips=15
Cost ~50 btc
Power= 450 watt

Profitable yes.
But consider that GPUs are still profitable, coz we havent hit the ASIC Wall yet. The recent 10-12% jumps might be small compared to whats coming. Everybody ignores bfl,in diff calcs and rightly so. But they sent samples to journalists. So they do have a working product even if it is an engineering sample. At some time in the future when the production ramps up,  all those pending orders coming into play, i cant estimate exactly how many.Benson might have an idea.

Using current available info,as of today,  till September(Avalon chips ship around then) nothing other than ASICminer is expected to ship.
(Do remember that you can sell your ASIC after breakeven at whatever Ghps/$ rate the other miners in September offer for a bit of additional profit)
Essentially, ASICminer has Free reign on Bitcoin kingdom and its humble subjects till September.

[Dashing]

Thanks Benson looks good

Off-Topic (Somewhat related),
I don't get it why people buy the 10 Ghash/sec miner from ASICminer (?) It's around 50 BTC each and well not so efficient either (130 watts?), Hopefully Avalon ships their chips soon and a few people can start to sell their own made ASIC miners with those chips

Timing Dashing. Cannot get mass Hashrates from anything else at this point in time.

 Yes, that's pretty true, it's the best hash rate producing device available at the moment

but the odd thing I've noticed is the initial ROI time for the mining blade from asicminer is the same as the time needed for the ROI for a small Lancelot FPGA, around 4 months. Which, is kinda bad for an ASIC, but yeah after the initial 4 months you'd profit pretty well I guess

But it isnt scalable.
A Lancelot costs, what, around 3.5btc including shipping.  
400 megahashes @30watt . It takes 4 months  to make 3.5 btc. For the next 3.5 btc more.
Considering a 3.5 btc investment its ok.
If you try to scale it to 10Gh/s,
Chips=15
Cost ~50 btc
Power= 450 watt

Profitable yes.
But consider that GPUs are still profitable, coz we havent hit the ASIC Wall yet. The recent 10-12% jumps might be small compared to whats coming. Everybody ignores bfl,in diff calcs and rightly so. But they sent samples to journalists. So they do have a working product even if it is an engineering sample. At some time in the future when the production ramps up,  all those pending orders coming into play, i cant estimate exactly how many.Benson might have an idea.

Using current available info,as of today,  till September(Avalon chips ship around then) nothing other than ASICminer is expected to ship.
(Do remember that you can sell your ASIC after breakeven at whatever Ghps/$ rate the other miners in September offer for a bit of additional profit)
Essentially, ASICminer has Free reign on Bitcoin kingdom and its humble subjects till September.

Yep that's true and yeah like I said,

after the initial 4 months you'd profit pretty well I guess

So heh, hopefully BFL ship soon

GPUs are profitable depending on what you pay for power

[laserwolf]

https://i.imgur.com/oGfCpzZ.png

This is a plot of Difficulty and a 2-month moving average of MtGox USD over time.
As you can see, they move in sync. The covariance is .98 (1.00 is perfectly in sync)

Which makes sense. The higher the cost of BTC, the higher the return from mining. Which increases the compute power entering mining. which increases the Difficulty.
When BTC drops, mining returns drop, compute power exits mining, and Difficulty reduces.

So Benson, when you use Difficulty increasing 10% per jump, there's a underlying assumption of BTC price rising around 10% every couple of weeks.

I think this is too narrow a range to consider. You should consider at least 3 price regimes:

• BTC stays about the same price it is now, say increasing only about 5% in the next 6 months.
  Closing at about $135 in December 2013. Difficulty will be about 12,000,000
  
• BTC goes through the roof. Closing at about $500 in December 2013. Difficulty about 50,000,000
• BTC crashes. Closing at about $20 in December 2013. Difficulty will drop to about 5,000,000

Just my 2 satoshi...

[IEA]

https://i.imgur.com/oGfCpzZ.png

This is a plot of Difficulty and a 2-month moving average of MtGox USD over time.
As you can see, they move in sync. The covariance is .98 (1.00 is perfectly in sync)

Which makes sense. The higher the cost of BTC, the higher the return from mining. Which increases the compute power entering mining. which increases the Difficulty.
When BTC drops, mining returns drop, compute power exits mining, and Difficulty reduces.

So Benson, when you use Difficulty increasing 10% per jump, there's a underlying assumption of BTC price rising around 10% every couple of weeks.

I think this is too narrow a range to consider. You should consider at least 3 price regimes:

• BTC stays about the same price it is now, say increasing only about 5% in the next 6 months.
  Closing at about $135 in December 2013. Difficulty will be about 12,000,000
  
• BTC goes through the roof. Closing at about $500 in December 2013. Difficulty about 50,000,000
• BTC crashes. Closing at about $20 in December 2013. Difficulty will drop to about 5,000,000

Just my 2 satoshi...

very nice, thanks

[nagatlakshmi]

What type of 'Jumps'?

[Benson Samuel]

So Benson, when you use Difficulty increasing 10% per jump, there's a underlying assumption of BTC price rising around 10% every couple of weeks.

Its a dynamic field. The 10% can be changed to reflect whatever the simulation requires.

[laserwolf]

check out this calculator:

http://www.coinish.com/calc/

[americandesi]

Thanks Benson looks good

Off-Topic (Somewhat related),
I don't get it why people buy the 10 Ghash/sec miner from ASICminer (?) It's around 50 BTC each and well not so efficient either (130 watts?), Hopefully Avalon ships their chips soon and a few people can start to sell their own made ASIC miners with those chips

Timing Dashing. Cannot get mass Hashrates from anything else at this point in time.

 Yes, that's pretty true, it's the best hash rate producing device available at the moment

but the odd thing I've noticed is the initial ROI time for the mining blade from asicminer is the same as the time needed for the ROI for a small Lancelot FPGA, around 4 months. Which, is kinda bad for an ASIC, but yeah after the initial 4 months you'd profit pretty well I guess

But it isnt scalable.
A Lancelot costs, what, around 3.5btc including shipping.  
400 megahashes @30watt . It takes 4 months  to make 3.5 btc. For the next 3.5 btc more.
Considering a 3.5 btc investment its ok.
If you try to scale it to 10Gh/s,
Chips=15
Cost ~50 btc
Power= 450 watt

Profitable yes.
But consider that GPUs are still profitable, coz we havent hit the ASIC Wall yet. The recent 10-12% jumps might be small compared to whats coming. Everybody ignores bfl,in diff calcs and rightly so. But they sent samples to journalists. So they do have a working product even if it is an engineering sample. At some time in the future when the production ramps up,  all those pending orders coming into play, i cant estimate exactly how many.Benson might have an idea.

Using current available info,as of today,  till September(Avalon chips ship around then) nothing other than ASICminer is expected to ship.
(Do remember that you can sell your ASIC after breakeven at whatever Ghps/$ rate the other miners in September offer for a bit of additional profit)
Essentially, ASICminer has Free reign on Bitcoin kingdom and its humble subjects till September.

Yep that's true and yeah like I said,

after the initial 4 months you'd profit pretty well I guess

So heh, hopefully BFL ship soon

GPUs are profitable depending on what you pay for power

Again, scalability to get such a huge hash rate is the problem with FPGA.

Secondly, Even if BFL starts shipping (which they have mentioned that they will start shipping from this week), except people in USA, others are going to be having a tough time with Customs.! Probably, it might get stuck there for atleast 10 - 15 days. Rumors are atleast 2 - 3 weeks.!

One of the good things of ASIC blades are that they come without a PSU and a case which makes it easier for them to pass the customs. They are being invoiced at $ 39.00 and sent as test boards. Still i have seen some boards stuck with customs in Europe.

Both my blades were cleared by U S customs within 12 hours. Otherwise do you think U S customs will clear a electronic equipment within 12 hours.? No way.!