Calculator:http://bitcoinx.com/profit/index.php
Being very generousWith a constant difficulty of 3438908 (unlikely) and a block reward of 50 BTC (highly unlikely), today's current BTC value of 12.12 (anyone's guess), hash rate of 832 (
https://en.bitcoin.it/wiki/Mining_hardware_comparison#FPGAs), a power draw of 80 W, a 0.10 $/kWH (probably average but most veteran miners could get slightly better rates) and a cost of $500, it will take
182 days before the unit breaks even.
But bar an act of god (or the devs), the block reward will halve sometime in the next day or so, so let's adjust our calculation.
Being somewhat generousWith a block reward of 25 BTC and the rest of the above being the same, it will take
1 year and 25 days to break even.
But this assumes that during that first year, the difficulty won't budge, which I'm sure no one here believes. Based on pre-order estimates, it's estimated the difficulty will increase any where from 7-40 fold. Please let's now assume that the difficulty doubles tomorrow and stays that way for the duration of our mining period
Being generousWith a block reward of 25 BTC, and the difficulty at 6877816 the duration of our mining period, it will take
2 years and 187 days to make even a single penny. Of course this all assumes that your electricity rate stays constant and the unit doesn't break or malfunction.
More realisticWith a constant difficulty of 34389080 (10 fold the current difficulty what I believe to be a reasonable estimate) and a block reward of 25 BTC, today's current BTC value of 12.12, hash rate of 832, a power draw of 80 W, 0.10 $/kWH and a cost of $500, it will take
never. It's so far into the future it's not even worth calculating.
But wait! The saving grace! The price of a BTC!!! Fair enough. I think throwing darts on a Mt Gox Live chart will have just as much luck as any of our predictions on the price of BTC, but let's use a quick 1,2,5 scale so that roughly x1 is not generous at all, x2 is generous and x5 is quite generous indeed! (Again, back of the envelope calculations and quite possible they are all lowballs).
Using the "More realistic" scenario with $10 a BTC, obviously the timeframe is still never.
With BTC going for $20 a piece, our pay of period is a shy under 27 years!
Finally, with one shiny Bitcoin selling at 50 Federal Reserve-funded Greenbacks, the payoff period is 3 years and 106 days!
Now here is where I think there is a glimmer of hope. With a mining period that long, it's entirely feasible that the price of BTC will have risen that high, what with decreased growth rate and probably more adoption. That being said, there a probably a half dozen other investments offering better mining returns. Unless the price drops to the $100, I think I will pass.