They are choosing a less risky option. By selling now, they get the price of the machine no matter what happens to the BitCoin economy. They are defining their core competency as being in the hardware business not the cryptocurrency business.

The same could be said for people selling pickaxes during the Gold Rush. Why not use the pickaxes to mine gold? They got the money for the pickaxes up front, then they were out of the picture. They didn't have to choose a place to prospect and worry about how long the gold would last. They chose to be in the pickaxe business not the mining business. Incidentally, during the Gold Rush it was these peripheral entrepreneurs that made much more than your typical gold miner.

I agree with RentFree's answer, but another part of the answer is the cost of capital. ButterflyLabs did not have the cash on hand to build its ASIC design or to manufacture all of the rigs. That's why they had their pre-order process: collecting the money up front was the only way they could afford to build the rigs. (Once they had built them, of course, they could have defaulted on their promises and used the units to mine for themselves instead of delivering them. But that would have had long term consequences to their reputation. I don't think the question was really asking if they should commit fraud, just whether they should keep their design private.)

The question does raise the question of whether it would be possible for a different company other than BFL (say a private equity firm with deep pockets) might fund their own ASIC mining rig and not release it to the public. I wouldn't say it's impossible, but it does have several factors against it. Firstly, RentFree's answer: that a private ASIC design strategy would have a lot of inherent risk. But there's also the issue that much of the cost of building an ASIC rig is in the design of the chip. Once our hypothetic private equity firm has spent the money to design the chip, the margin on individual rigs would be fairly high. So, at that point, there would be little incentive for them to refuse to sell the rigs to others. (Even if it raised the mining difficulty for their private rigs.)

In short, while there is nothing to say the ButterflyLabs (or other ASIC manufacturers) might not use some of their rigs (or test rigs) for mining, the business model of these types of companies wouldn't work very well if they didn't sell the majority of their machines: the upfront capital costs are too high and the incremental profit margin is too tempting.

I Would not be surprised if they collect cash up front for many orders, create a production run which takes 30 days, mine the units for 30 days, and then deliver. That's how I would maximize profit. They take no risk, but having the consumers pay for the costs (and more), they then have many many machines to mine or "perform QA (Quality Analysis) to make sure they work correctly while profiting them. After that, they package the units and send them to the consumer. If there's a spike it bitcoin value, I'm guessing the "QA" takes longer while mining is more profitable.

Just a speculation, but being a business owner, and having run production runs and QA through factories all over the world, this just makes sense to me.

I believe it's primarily because at such a large leap forward the bitcoin network would fail. I've read for starters that if someone owned over 50% of the hashing power the network would be compromised, I don't know this to be true or otherwise, but what I do know is, Bitcoins have little value if no one is trading them because their value is based on trust. If one company drove up the difficulty and also had a large majority of the coins out there then people would lose interest in the bitcoin network and therefore they'd not have all that much value either another network would ultimately just replace it.

Gather too much hashing power in your hands, and you compromise the security of the Bitcoin network. If security of the network is compromised, the price of bitcoin falls through the floor, and no matter how much you can mine (or manipulate the network into giving you) you end up with millions of worthless bitcoins.

It's much smarter to distribute the hashing power between the users, keeping bitcoin price high, and earn cash for the product. With Bitcoin, as opposed to normal capitalist markets, monopoly is disastrous to the whole market including the monopoly owner, so any enterprise aware of approaching the status of monopoly will actively fight reaching it - for example, selling a part of their processing power.

They do mine. Hashrates tripled during the last four months, and so did the difficulty. While this tendency continues, the difficulty will keep on rising. When you get your ASIC device from them, you are only able to mine $10 with two Jalapenos in one month. No sooner, no later.

If you watch the video on their website, there is a server room with racks and racks of networked miners so its possible that they are. However, it also appears you can purchase processing power from them instead of the actual hardware. The network setup might be for that.

It's totally possible that they have used hardware which they sold to customers for their own mining operation. Your gear may very well have gone to a warehouse somewhere for "testing". Once there's very little value left they ship it.

It would be unethical and likely illegal but by far the most profitable thing to do in their position. I would challenge BFL to prove they haven't done this by making their factory shipping manifests and bills of lading public.

Just a thought. I'm not (yet) a miner but seen all the complaints in delivery time from BFL it makes me wonder if they really don't do some mining with the production units before delivering them. It seems normal a 6 month delivery time wait, what doesn't happen in any other kind of market... Would you wait 6 months to by a standard electronic device?

That's the dark side of this business, and the main reason I don't get decided to invest here...

Consider also that BFL may be using the initial pre-order capital to invest in and use infrastructure prior to actually shipping. Unless you are purchasing one of the really, really big rigs, by the time you get your ASIC miner, it will be relatively worthless. Similar to what a cpumining is.

This is really a type of ponzi scheme. Before you get angry at me for saying it, really think about it...

Soon, bitcoin mining will only be really profitable for those who really have infrastructure... or for the mildly curious who don't care about making profit.