Hashlet Genesis, which is apparently due for a price drop. If you follow that link right now, it's still showing the original price of $9.95 per 10GH, but on the ZenMiner Portal (registration required) the price has dropped to $7.95 per 10GH.
This is hardly surprising considering the initial ROI estimate at $9.95 was previously four months, more likely six after difficulty adjustments. The price of Bitcoin has currently dropped 10% in the past couple of weeks, however, and with difficulty still increasing that means the forecast for Genesis got worse. The price cut is thus not just a good thing to offer customers; it's pretty much required if GAW hopes to sell any more Genesis Hashlets -- and we're probably due for a drop in the maintenance fees to $0.01 as well. Here's why:
Let's start with a worst-case scenario. If the Genesis continues to sell at $7.95 per 10GH and has a $0.02 per day maintenance fee, with a 15% rate of difficulty increase it would become unprofitable to run in about four months, and even at $7.95 you'd still lose $5 or more on the purchase. So this is a losing proposition.
Next up, let's go optimistic and use 5% difficulty increases. A $7.95 Genesis actually earns money this way, but just barely -- you'd get a profit of nearly $0.05 in about 11 months. Oops, that's not good either. Can you see why Genesis isn't looking too hot right now? Even at $0.01 per day per Genesis, you'd still need over six months to break even and you'd be down to earning pennies a month of "profit" -- with a final net income of $4.60 in about 18 months.
Okay, try this on for size: let's say it was free to run the Genesis Hashlets as a best-case option. It still would require four months to hit ROI with a 5% difficulty adjustment, or over six months at 10%. Either Bitcoin mining hardware is going to need to get a whole lot more efficient, or we're quickly approaching the point where Bitcoin difficulty reaches a maximum value and either or holds steady or starts dropping. Or maybe price jumps up to $1000+ again and everyone goes crazy.
The bottom line is that Bitcoin ASICs are becoming a highly questionable proposition for miners, and even manufacturers are likely starting to worry about margins. No one sane is going to take a risk on thousands of dollars of pre-order hardware anymore, and really they shouldn't have to. We're also fast approaching the point where more powerful ASICs become unlikely. We'll probably see 10TH ASICs next year that use 1000W, and maybe even 20TH ASICs that use 1500W, but we're not likely to see triple digit TH rates from a single piece of mining hardware any time soon.
Update: GAW has apparently decided not to drop the price of the Genesis Hashlet, at least not at this time. ("It was a mistake. Oops, sorry.") So, umm... don't buy any unless you have faith that GAW will do something else to make these worthwhile in the long run.