Last year average increase every two weeks was 3,74%. This year it is already 12,87%. And it will go up after all Avalons will be delivered. And then there will be bunch of DIY miners from Avalon chips, and probably BFL. Very soon mining will be profitable only for miners with rare elite hardware.
Which is why I chose 20% to illustrate my example, it may seem high now but once the next batch of Avalons and all those DIY chip miners come on stream it can ony go upwards. Perhaps the chip purchases will be profitable. Avalon is roughly 0.08
BTC for 300MHash/s plus assembly costs, but you're still looking at up to 3 months delivery on new orders - 10 weeks for the chips, then all the business of getting them built onto mining boards. I'd have taken a punt if I had the
BTC in hand, but the minimum order is typically 10 chips, and I'm a bit light on coin at the moment.
I will definitely not buy chips. For not engineer it seems impossible to assemble something useful with them
I am planning to spent few thousands on working ASIC. Avalon batch 4, or BFL if they will prove their legitimacy, or other new dealers on market.
But, again back to difficulty. Its impossible for it to grow geometrically too long.
With 120% increase of difficulty every two weeks for next 6 months - result will be around 100,000,000.
This number seems impossible for me. The thing is that each ASIC does not increase difficulty by multiplying it to some coefficient. Otherwise, it just adds some value to it. And total amount of added difficulty is limited by ASIC production speed. Current difficulty is 12,153,411. All Avalon batches #2 and #3 ASICs together are 1200 pieces * 70GH/s each = 84TH/s. This will give difficulty increase for 12,000,000. If some amount of them is already delivered and working - increase will be even less.
Then, there will be BFL. I have no idea about their produced ASIC quantity. Lets assume BFL will provide total hashing power the same as all Avalon ASICs together - 1500 pieces * 70GH/s each = 105 TH/s (corresponding difficulty increase is 15,000,000).
Totally that will be 12,153,411 + 12,000,000 + 15,000,000 = 39,153,411 (difficulties).
To fill remaining 60,000,000 difficulty points - someone will need to produce chips with total hashing power of 430 TH/s (it is Avalon chips, BlockEruptors, private ASICminers, BFL chips..). To compare - current power of all miners in network is around 86 TH/s, and is increasing for ~8 TH/s every two weeks. Additional 430 TH/s is way to much, it seems impossible.
I'd say that difficulty will reach this 40M mark as soon as Avalon and BFL will deliver ASICs, but then it will grow much slower.
Now in my calculations to find possible profit of ASIC - I assume 2M increase of difficulty every two weeks (two times bigger than current).
In this case - BFL 50GH/s received in September will generate 270 BTC profit in two years.
USB Block Eruptor received today will generate 2,5 BTC in 2 years (and then subtract its price of 2 BTC and you have profit of 0,5 BTC. In case if difficulty increase is only 1M/2weeks (like average now) - then profit is bigger for 1 BTC. But its best case scenario, better profit is impossible. If difficulty increase will remain 1M/2weeks - it will be completely unprofitable to buy it after 6 months. In reality it will become unprofitable even faster)
