I was speaking hypothetically in the realm of "if BFL was able to ship right now in volume"
This would rely on the assumption of two things: that they could fill their existing orders of $125/$275 quickly, and that every new order placed could be priced at roughly 14-15x the current price - and people would pay it (following that people paid that much per mh/s on the USB miners)
It's very simple to logically believe that IF BFL could ship right now, they would, because they would have a monopoly on ASIC sales given their hash rates... and they could charge a LOT more for each unit.
The money they could make in BTC a day would be nothing compared to the $ they could be bringing in and then just converting to BTC
Sure, assuming they could price their product 15x and ship in bulk right now, then they would be better off shipping product. However, I don't believe they can do either.
The ASICMiner product is not really competitive. Even so, it and the Avalon chip sails have already sucked up a lot of capital that would have otherwise gone to BFL. Also, both products are being sold at inflated prices. If BFL actually enters the market, Avalon and ASICMiner could revise their prices down to compete.
But they're not capable of or willing to ship in quantity either. A couple hundred units from Avalon? So what? They aren't capable or or interested in shipping in quantity.
1 Avalon unit is 66GH/s, more than 10x the performance of 1 BFL Jalapeno (the only BFL unit that has left their facility). Each Avalon unit has 3x88 chips in it (IIRC) so that is (300 batch 1 units + 300 batch 2 units ) x (3 x 88) = 158,400 chips. That is volume. Also, they have sold another 400,000+ chips to the DIY and group buy crowd. That is shipping ASICs in volume. BFL by contrast has moved maybe 3 dozen ASIC chips.
And neither is ASIC Miner. They've said several times that the only reason they're actually selling at all, is that they can't bring more equipment online until there's mor hashing power out there, to prevent 51% of network share. So they're only selling in the mean time. Once BFL comes online, full-force, Avalon is toast, because they can't produce at speed or quantity, and ASIC Miner are gonners too, because they don't even want to be in the retail game in the first place.
BFL miners compete with both ASICMiner retail and ASICMiner mining operation. So the ROI on BFL units does not distinguish between ASICMiner chips.
If BFL starts shipping in quantity, difficulty is going to SKYROCKET, and that will create even MORE demand for their product. Anybody who spends more than a few cursory seconds on this realizes that there's WAY MORE money in selling the 'shovels' than there is in mining.
Guaranteed, there's nobody who wants BFL up and producing (and shipping) at maximum capacity more than BFL does.
The facts do not bear out your statement. $2700 per day per 500GH/s for mining in the short term vs $5000 to $10000 profit per 500GH/s selling. It makes more sense to mine first, then ship. Only after the difficulty rises or the price of USD/BTC falls will the tide shift. Don't forget, BFL can mine and ship, they do not have to choose one or the other.
Per chip "quantity" isn't relevant to tapping the market, because the majority of the market doesn't need or want an $8000 mining machine that consumes 5 times as much power and costs 2-4 times as much to purchase.
They don't use 5 time as much power. They use 1.5-2x as much. (13 x 50w ars technica @ 5GH/s vs 600W)
https://en.bitcoin.it/wiki/Avalon#Power.
The first batch was $1500 per 66GH/s and they made money off it. The second batch is roughly $8000 but that was an anomalous result because Avalon explicitly said they were building into the cost of the unit the cost of upgrading their production line. If BFL ships, they could conceivably lower their cost back to $1500 and still make money and have a better price point than BFL. Presuming of course they don't have to buy a new factory every time they do a new run.
ROI on BFL units does not distringuish between ASICMiner chips? What does that even mean? ASICMINER will NOT be selling units, as soon as they've got some competition. Meaning the ONLY "ROI" that will be factored will be the buyers of BFL (and a few Avalon) units.
Since the ASICMiner mining operation holds roughly 25% of the total bitcoin network hashrate, any buyer of BFL is competing with them. They have the capacity to expand this more but they don't want to hold too large of a percentage of the overall network because it could destabilize bitcoin itself. Every addition to the network by someone else allows ASICMiner to bring more capacity online. The higher the network hashrate, the lower the ROI is on a BFL unit. Therefore it does not matter whether an ASICMiner chip is running in the ASICMiner facility or in someone's home because they bought it retail. In either case each ASICMiner chip lowers the overall ROI of BFL units.
What do you mean it makes more sense to mine first and then ship? Mine with how much, for how long? If they're mining, where is the hashing power hiding out? Where's your evidence that they are (or even intend to) mine with their customers' products before shipping them?
First they mine for some period of time (1 day or 1 week or 1 month), then they clean off the unit with compressed air, box it up, and ship it. As I have said numerous times, BFL does not have the capacity at the moment (perhaps ever) to be visible in the network hash rate. They could easily add 500GH/s and nobody would notice if they used multiple pools.
My assertion is that if they had 200ish chips (~500GH/s) it would make far more sense to mine with it than to ship out the units. They cannot clear their order book with 100 Jalapenos (or get anywhere close to it). They could easily disguise the addition of 500GH/s (0.5% of the total network). If they did that for a month it would earn $80K at current prices. That would be a powerful temptation to a company that is in a cash crunch.
I have no evidence that BFL is actually mining, I am only asserting that it would make economic sense for them to do so.
In fact, I doubt they have anywhere near 100 units operational at this time (also conjecture on my part).
