Some research will provide more useful information.

Every 2016 blocks, the Bitcoin network determines how long it took to solve those 2016 blocks. If it took LESS than 14 days, it INCREASES difficulty. if it took LONGER than 14 days, difficulty is decreased. The Network tries to keep the network solve rate right at 2016 blocks in 14 days. Difficulty increases happen way more often than decreases. To get a little history:

https://bitcoinwisdom.com/bitcoin/difficulty

Scroll down a bit, and there is a list of the last sereval adjustments and when they happened, and by how much. I personally find the calculator at the top of the above page to be very useful.

Keep hoping........

Bitmain has approximately zero reason to extend the warranty or otherwise improve their terms. Maybe they'll reconsider if/when a batch takes a week to sell out?

I'll also add that "Hardware" Sub-Forum has some useful threads on the SP20. Th following is just one of them:

https://bitcointalk.org/index.php?topic=872014.0

I won't weigh in on your specific choice of PSU, but I will tell you that you will almost certainly be much happier if you target your SP20 for 1200GH or perhaps a bit lower. It's unlikely you'll ever get to the fabled "1700 GH", and the closer you get the hotter, noisier, and less efficient it will run. Your running costs will be lower both in absolute number and in terms of $/GH at the lower speed. Most past owners of SP20 hardware really think it should have been advertised as a 1400 GH miner with possible 1700 GH over-clock potential. With reduced speed also comes reduced heat, and reduced noise from a slower turning fan. The demands on your power supply decline as well.

I ran my SP20 at about 1000GH using a 1000W EVGA power supply, with power to spare on the PSU/wiring, and reduced heat and noise (the biggest benefit for me).

I think that this is virtually un-knowable. I am sure it will be much higher. In twelve months I expect it will be at least 30% higher than it is today. A higher BTC price always ramps up the hashrate as people buy more gear, and "marginal" gear continues to be run.

Just my $.02 on the subject.

With the increase in BTC price, you can absolutely count on difficulty going up. When BTC price rises, people add hash rate one way or another. When that happens, difficulty will increase. Over the course of 5 months, it will be at least a 20% increase (IMHO).

While I don't own an R4, you might be able to "level out" the temps by standing the R4 in it's side, assuming that's possible. I did something similar with an SP20, though it was a more square shape at the end  end and easy to do.

It's only "very profitable" if you ignore the initial cost of the solar installation, and have no other way to use the electric power that is generated. If you have any other electric load nearby (say a residence), and you are paying actual money for that load, then it is in fact costing you something to mine.

For an economist, I think that's called "opportunity cost".

Do you have consistent, reliable Internet access? What about temperature and other environmental?

Disposing of IED's sounds very remote and possibly inhospitable. It also sounds like somebody might eventually care about electricity consumption of any serious scale.

I think we can let this thread dies a natural death since we are approaching 6 months after the halving, and difficulty is more than 286 Billion.

Just to be picky, time doesn't figure into the ability to generate 1 BTC/month. Great electric rates, and luck do factor in.

Just to be clear, given your definition of "100% return on investment", if you stuffed your money in a mattress and it didn't burn up or get stolen, you'd have "100% return on your investment" (i.e. you just broke even).

Anybody with $100K to invest should understand that getting you money back is 0% ROI, the only profit you would have would be the residual value of the miners. If you think they are still valued at 100% of their original cost, THEN you would have 100% ROI.

Too bad it costs $999....... 

We really should let this thread return to a quiet state.

With all due respect, I think you are forgetting the current position of Bitmain. At the present time, they are a near monopoly on miners of this efficiency. If you don't like what they are charging, you are free to not buy from them. Unless and until there is a near alternative, they can get away with a lot that might otherwise be unacceptable.

I would assume that reference is to some flavor of Altcoin that is then exchanged for Bitcoin. The Altcoin is presumably easier to mine, and the exchange good enough to be more profitable in terms of electricity and initial investment.

Right now Bitmain clearly holds "monopoly status" on efficient Bitcoin mining hardware. That situation is NOT true on most any other Altcoin today. GPU mining is open territory for the most part, and complete usable on the next Altcoin that may come along.

It always strikes as odd that this works. It feels kind like mining copper and then exchanging it for Gold since you don't trust the Copper market. I always wonder who is on the other side of the exchange. What person want to trade a BTC for ETH or some other coin?

So the connector that plugs into the Raspberry Pi is actually 40-pins?

If so, then they must have developed their flash image on a Model B+. As I recall, MOST of the first 26-pins of the GPIO map directly onto the 40-pin GPIO, but there a couple of small differences that don't matter to many applications.

While I know approximately zero about the KNC hardware, the very Original Raspberry Pi Model B (512 MB), had a 26 pin GPIO header. The subsequent Model B+ that are shown above has a 40-pin GPIO header. My recollection is that the Pi microcode, NOT the stuff supplied by KNC, had to be changed to support the Model B+ and it's expanded GPIO [pins.

So what king of connector is on the KNC hardware? If it's a 26-pin variety, in which case the microcode that's included within their flash image is likely meant for just the original Model B, and may not support the Model B+ (nor the Pi 2 or Pi3).

My apologies if this obvious to everyone else.

You omitted the ^TM, which I rightfully think belongs to Butterfly Labs....

In case it isn't obvious already, you should EXPECT difficulty to increase each month over the long term. Exactly how much is unknown. At the end of a year, you could easily be down to say .7 BTC/month.

You also need to ask yourself if you have the right environment (electricity, heat, noise tolerance) to operate the required equipment to mine 1 BTC/month.

Do some research onto the Bitcoin protocol, and how it behaves long term before deciding on what you are getting into. While 1 BTC/month sounds like a lot of money, it isn't easy or cheap to produce on a long term basis.

Short Answer: NO

Long Answer: While as efficient as the S9 may be, $.08 Electricity cost is just too high for comfortable, profitable, Bitcoin mining these days. Over the course of a year, you can expect difficulty to increase at least 25% (probably more). If BTC price increases that's great, but as BTC price increases, so does hahsrate and then difficulty.

One hundred S9 miners is a lot to manage, and keep functional if I am to believe what I read. Quite frankly I think I am better paid at my "Day Job".

Just my $.02.