Should you Cancel your Jalapeno Order? - This might help to decide

[Swimmer63]

Here is my estimate on revenue per week for a Jala 5Gh/s and 7Gh/s.  Assumptions: $111USD/BTC, 19% straightline increase in difficulty every two weeks.  I actually used 9.5% per week.  No allowance for electricity usage, just the gross $.
A 5Gh/s will make about $1000 between today and the end of the year.  Next year you only have a few weeks before it costs more to run than it makes.  So it's a brick by the end of February.  Although it may have a second life with altcoins.

Week   5 Gh/s   7 Gh/s
19-Jun    $101.27     $141.78
26-Jun    $92.48     $129.48
3-Jul            $84.46     $118.24
10-Jul    $77.13     $107.99
17-Jul    $70.44     $98.62
24-Jul    $64.33     $90.06
31-Jul    $58.75     $82.25
7-Aug    $53.65     $75.26
14-Aug    $47.10     $68.60
21-Aug    $43.01     $62.64
28-Aug    $39.28     $57.21
4-Sep    $37.32     $52.25
11-Sep    $34.08     $47.71
18-Sep    $31.13     $43.58
25-Sep    $28.43     $39.79
2-Oct    $25.96     $36.34
9-Oct    $23.71     $33.19
16-Oct    $21.65     $30.31
23-Oct    $19.77     $27.68
30-Oct    $18.06     $25.28
6-Nov    $16.49     $23.09
13-Nov    $15.06     $21.08
20-Nov    $13.75     $19.25
27-Nov    $12.56     $17.58
4-Dec    $11.47     $16.06
11-Dec    $10.48     $14.67
18-Dec    $9.57     $13.39
25-Dec    $8.74     $12.23

[1714996815]

1714996815

[1714996815]

1714996815

[btceic]

Just as a quick FYI,
People that ordered a year ago are just now getting their orders.

[Gomeler]

Just as a quick FYI,
People that ordered a year ago are just now getting their orders.

And I think they are shipping an an accelerated pace versus pre-order date. I know a week ago or so they shipped a few weeks worth of jalapeno's in one day.

[Swimmer63]

If you get it before September you will make some money.  But the window is small.

[shmadz]

<snip>
Next year you only have a few weeks before it costs more to run than it makes.  So it's a brick by the end of February. 
</snip>

I'm a bit confused. How much does a jalapeno actually cost to run? I thought they only pull around 35 watts?

I don't know what electricity prices are but isn't it around $.20 per kwh? or lower?

that would make the cost of running a jally for a week around ... 0.035kw * 24h * 7d * $0.20/kwh = $1.176 per week?

Unless my math is wrong, which is probably the case...

Could you run these numbers for say, 1 year out? until july 2014? I'm genuinely curious.

[hl5460]

Here is my estimate on revenue per week for a Jala 5Gh/s and 7Gh/s.  Assumptions: $111USD/BTC, 19% straightline increase in difficulty every two weeks.  I actually used 9.5% per week.  No allowance for electricity usage, just the gross $.
A 5Gh/s will make about $1000 between today and the end of the year.  Next year you only have a few weeks before it costs more to run than it makes.  So it's a brick by the end of February.  Although it may have a second life with altcoins.

Week   5 Gh/s   7 Gh/s
19-Jun    $101.27     $141.78
26-Jun    $92.48     $129.48
3-Jul            $84.46     $118.24
10-Jul    $77.13     $107.99
17-Jul    $70.44     $98.62
24-Jul    $64.33     $90.06
31-Jul    $58.75     $82.25
7-Aug    $53.65     $75.26
14-Aug    $47.10     $68.60
21-Aug    $43.01     $62.64
28-Aug    $39.28     $57.21
4-Sep    $37.32     $52.25
11-Sep    $34.08     $47.71
18-Sep    $31.13     $43.58
25-Sep    $28.43     $39.79
2-Oct    $25.96     $36.34
9-Oct    $23.71     $33.19
16-Oct    $21.65     $30.31
23-Oct    $19.77     $27.68
30-Oct    $18.06     $25.28
6-Nov    $16.49     $23.09
13-Nov    $15.06     $21.08
20-Nov    $13.75     $19.25
27-Nov    $12.56     $17.58
4-Dec    $11.47     $16.06
11-Dec    $10.48     $14.67
18-Dec    $9.57     $13.39
25-Dec    $8.74     $12.23

Asic could be used to mine PPC. With recent increase in PPC diff, I am expecting to mine PPC with my pre-ordered avalon. Hopefull PPC will triple the price on its 1st anniversary.

[logicbomb666]

A 5Gh/s will make about $1000 between today and the end of the year.  

I would love to think this is true. :-D

[Swimmer63]

@hl5460 - that is definitely an alternative.  There are other sha256 coins and often they are more profitable to mine right now.
I am guessing that by the end of the year we will see more sha256 coin debuts as anyone with less than a 20-30 GH/s rig is going to be making pennies.

@shmadz - I will try to run the numbers today.  But they are not going to be pretty.

[DallasMiner]

Where I'm at, with electricity at about 0.085 USD per kwh, a Jalapeno will cost me about 7 cents a day to run. So, even when difficulty rises to the point that a Jala is mining .0014 BTC a day, it's still giving a 50% profit margin (at today's rates). That's really not too bad, in all honesty.

[Swimmer63]

True on the power / cost issue.  The thing I was trying to show though was how dramatically the revenue drops off over a short period of time.  What is giving you 0.0014 BTC one week is giving you 0.0009 the next week and so on.
Basically the window for this hashing rate (5GH/s) closes very quickly over the next 6 months.  If you look on Ebay some people are paying $2,000 for this and it will never come close to making that back, at least in BTC.  Unless of course something changes and the difficulty stops increasing at this rate.
I will finish the numbers until they get down to nothing.  Just been too busy this morning.

[Loredo]

Likely, pretty close to right.  One comment, though:

Some have begun using "network proportion" rather than difficulty in proformas.   That is, if the network at a point in time is T Thps, and you are h Thps, ("h" for jalapeno  ), then I'd proforma dollar daily performance as (h/T * 3600 * USD/BTC).  One advantage is that you can then continuously (or cyclically, or otherwise) "grow" network hash, whereas difficulty is stepwise, and in fact the instantaneous probability of you discovering bounty is a function of the network hash rate.

Using 150 Thps, $111 BTC, and h = 5/1000 Thps, I get $8993 for the week, in noise-band with your $101.

Just sayin'.

[Swimmer63]

OK -  I stretched the projection out for a year.  The weekly numbers are exact given the growth assumption.  The monthly numbers are approximate.  For ease of reading I made it a Google Doc file. 

Conclusion is the same really.  While the rate of drop off in revenue decreases the higher the difficulty rises, you are not making much by then (January) so it's not much of a help.

If your Jala cost you $300 total then anything you mine prior to September 25 is profit.  Because after 9/25 you will only make about $300.

I did this just for me to figure out whether I should cancel my orders.  I have not decided.  But I am definitely not going for the upgrade to 7 GH/s.  Then you are stuck with it as in no refund.  I am thinking if we get to August and still don't have it in hand I might as well cancel.

Of course there is the opportunity to mine other sha256 coins.  That may change the decision.

[Swimmer63]

Likely, pretty close to right.  One comment, though:

Some have begun using "network proportion" rather than difficulty in proformas.   That is, if the network at a point in time is T Thps, and you are h Thps, ("h" for jalapeno  ), then I'd proforma dollar daily performance as (h/T * 3600 * USD/BTC).  One advantage is that you can then continuously (or cyclically, or otherwise) "grow" network hash, whereas difficulty is stepwise, and in fact the instantaneous probability of you discovering bounty is a function of the network hash rate.

Using 150 Thps, $111 BTC, and h = 5/1000 Thps, I get $8993 for the week, in noise-band with your $101.

Just sayin'.

I get it, which is good for me.  I am not the most mathematical.  What's the 3600 - coins per day??

[Loredo]

Likely, pretty close to right.  One comment, though:

Some have begun using "network proportion" rather than difficulty in proformas.   That is, if the network at a point in time is T Thps, and you are h Thps, ("h" for jalapeno  ), then I'd proforma dollar daily performance as (h/T * 3600 * USD/BTC).  One advantage is that you can then continuously (or cyclically, or otherwise) "grow" network hash, whereas difficulty is stepwise, and in fact the instantaneous probability of you discovering bounty is a function of the network hash rate.

Using 150 Thps, $111 BTC, and h = 5/1000 Thps, I get $8993 for the week, in noise-band with your $101.

Just sayin'.

I get it, which is good for me.  I am not the most mathematical.  What's the 3600 - coins per day??

Yep: (25 BTC / blocks * 6 Blocks / hr * 24 hr) = 3600 coins / day.

In complete generality, it should be (where b denotes blocks) (25 BTC / b * B(T,d) (b / h) * 24 hr ), and
B(T,d) is a deterministic function that maps hash rate T under difficulty d to the expected blocks per hour.  

If nominal blocks B_n is the solution, T_n, to B(T_n,d) = 6, then if T < T_n, expected blocks per hour falls, and vice versa.  This is what some guys were freaking out about yesterday when T plunged for some reason; network Bph dropped sharply, in the eyes of some, "dangerously" so.

But, IMO, the above level of complexity isn't needed for pro forma calculations, since the uncertainty around future net hash rate is far greater than these subtle, transient, largely self correcting, effects.  3600 coins per day is the normal value with pretty small variance over time.

EDIT: Network is now over 180, well above T_n currently, and mining is smokin' along at over 8 BlocksPerHour.

[k9quaint]

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

If you could pay $274 for a Jalapeno and receive it at today's difficulty, it would be a good deal.
Unfortunately, there is about 300TH of orders in front of you. So it is impossible to receive a Jalapeno ordered today at today's difficulty.

Try starting the calculation at 60,000,000 difficulty and then see how much you make.
Also, 1.3% per day is a very low estimate for how much the hash rate will grow in the coming months.

[cdog]

Ive been setting the difficulty to 100,000,000 I think its more realistic, but BFL could easily go under before anyone receives anything.

Thats the real issue, not ROI but total loss of investment...

[Swimmer63]

https://docs.google.com/file/d/0B_tqfkR9Qrx0WVdmNmQzU1FzcEU/edit?usp=sharing

Edit:  Fixed link

[suprabitz]

There's a point where the exponential growth will taper off. Because right now people are putting 60gh/s and 500gh/s units online and blowing out the competition running FPGA and GPUs at 10x slower speeds.
When the majority of these BFLs ship the majority of people who want Asics will have Asics. Yes there will be people who say I want MORE asics but that will be a slower growth than the jump from FPGA/GPU to Asics. It will be more linear as we saw with people who had GPUs and FPGAs ordering more of those... the rate will rise linearly and not exponentially.

When will that happen is the question... I suspect once production catches up to demand and you can go online and order an Asic and have it ship out within 2-3 days.

Incidently we saw the same growth from CPU to GPU mining, it grew exponentially and leveled off once the market was able to move to the newer hardware. Right now the bottleneck is GETTING the hardware so the few people who get them ahead of the rest, get to make 10x the profit.

[k9quaint]

There's a point where the exponential growth will taper off. Because right now people are putting 60gh/s and 500gh/s units online and blowing out the competition running FPGA and GPUs at 10x slower speeds.
When the majority of these BFLs ship the majority of people who want Asics will have Asics. Yes there will be people who say I want MORE asics but that will be a slower growth than the jump from FPGA/GPU to Asics. It will be more linear as we saw with people who had GPUs and FPGAs ordering more of those... the rate will rise linearly and not exponentially.

When will that happen is the question... I suspect once production catches up to demand and you can go online and order an Asic and have it ship out within 2-3 days.

Incidently we saw the same growth from CPU to GPU mining, it grew exponentially and leveled off once the market was able to move to the newer hardware. Right now the bottleneck is GETTING the hardware so the few people who get them ahead of the rest, get to make 10x the profit revenue.

A lot of people will not make a profit on the ASICs they bought. They would have been better off getting their BTC from Mt. Gox.

[KyrosKrane]

All the calculations I've seen about rate of return on Bitcoin mining tend to assume that $/BTC will remain stable. Is that a valid assumption?  I'd think that given the decrease in coins produced per person mining, that would effectively act as a reduced supply in economic terms, sending the dollar value of a Bitcoin upwards.