Your electricity cost assumption is VERY understated. Don't forget tiered electric rates which mining definitely breaches. In California you are looking more at $0.36/Kwh.
In the Bay Area you only get $0.13 rates for the first ~330Kwh per month which you will breach within 2 days. Rates then jump to $0.15/Kwh for the next 100Kwh and then $0.32/Kwh for the next ~230Kwh then $0.36/Kwh for the rest. Within a week you will be paying $0.36/Kwh. Triple the costs.
I went with what my electric cost is. Your value will of course vary. I'm locked in at 10.5c/kwh in winter, 11.5c/kwh summer. One of my partners is in California, and I hear him curse the costs out there all the time. He's starting to switch his farm to Scrypt this week. As we speak actually. We've negotiated a near-wholesale deal for Scrypt systems.
You actually don't even need scrypt-n ASICs to be profitable, mining Litecoin with enough ASICs can be already profitable enough, since you can have more than 10 times the hashrate/watt (2 gridseed equals roughly one 280x but with less than 30w). Means, that would bring me 50MH now with same powercost.
You don't need Scrypt to be profitable either. After covering electric costs, I'm still raking BTC in with my GPU farm, and enjoying the heat as well. My point was that mining for GPUs is only profitable if GPUC is significantly more valuable than it is now. Which might occur one day.
about 10k per 1mh/s
Thank you. So let's say 7.5k/1mh as another poster mentioned a similar figure based on difficulty. Let's speculate:
If I mine 75k GPUC in 24 hours, and GPUC is worth 0.00000115, then that's .25875btc (presuming btc = 670usd). If the store owner worked to get a deal as good as Amazon gets, he could sell a 270 for exactly that cost.
Now as a miner, it cost me $55 in electric to earn that. So in reality for the 225,000 GPU purchase, I actually need to mine 296,000 GPUC to cover all costs. What I could've done was simply trade 225k GPUC to BTC, and then BTC->Amazon, and get the card that way.
Totally circumvents the business frontend, which is required to sustain GPUC. (edit:) Which is why i said earlier, it would demand that the market value of GPUC still be far above what it is presently.
I can get MSI R9 270's on Amazon for $150 fiat (last week's order).
How??? MSI R9 270s are at least $250 on Amazon when I search and I haven't seen them below $200 since last year.
Last Monday Amazon got one of its weekly shipments of GPUS. MSI R9 270's were $171 Prime. Factor in my 15% discount (which anyone using BTC can get), the cards were $145. Add $4 for overnight delivery, and it's pretty much $150/card. With no quantity limits.
You have to be quick on the draw. I'm constantly refreshing Amazon inventory to see what's available. What usually gets leftover are the Amazon 2nd and 3rd party resellers, which can set the card at any price they want. $200.. $300.. etc.
(edit:) Let me put it this way, 3 weeks ago when MSI 270 came in stock at $190, Amazon received a truckload of them. Several hundred to say the least. In 6 minutes they sold out, and 1 minute after that they sold out of their backorder limit. 2 weeks prior to that I saw them sell out of 250 cards in 60 seconds. I was part of that.
You wrote too much... cal: 75k GPUC in 24 hours, and GPUC is worth 0.00000115, then that's .25875btc (presuming btc = 670usd)
75.000 x 0.00000115 = 0.08625btc (~53usd)
I've mined all first day and earned 7GPUC (before 2600 blocks), and a day after and got >20.000gpuc and ....... ~ 12usd for 2 days....
Thats enough for loving this coin and hopeless now... Nobody mine the coin with price lower than electricity cost ofcouse at least for 2 days... I quit and mine another coin today, if the price of gpuc stand and lower, i still love and buy cpuc... No mining anymore sure. What the DEVs doing for themself but we mine we have to pay the power we spent...