Great. Another iSuppli teardown that's going to cause everyone on the Internet to cry foul at how Amazon is charging 48% mark up on it's devices. Yes, because these things build themselves, weren't designed by anyone, get shipped through instantaneous teleportation, market themselves, and do their own support...
The only reason it isn't lower is because they can't make them fast enough.
I'm sure Amazon would sell these at a small loss if it meant grabbing marketshare. The most convenient way to get a book on the Kindle is through Amazon.
You are, at the very least; in the upper quartile of technical ability of Kindle owners. You underestimate what the typical user is able or interested in. It's not that they couldn't figure it out if motivated. But mostly just that they can't be bothered.
Cost is actually determined by both the demand and the supply. This is because Amazon hypothetically takes into account their marginal revenue (which comes from demand) in relation to the cost of producing it, in figuring out a price point. Specifically, producing up to the point where MR = Cost. (Sorry for being an econ snob)
This is good news. It means that, given enough competition, prices can fall pretty fast. Even better news is that the E-Ink display makes up most of the cost, so if the prices on those go down, it could make a big difference.
John Gruber wrote a great commentary on iSuppli's methods a while ago [1]. Basically, his assertion is that their numbers are a mix of generalizations and pure imagination, and don't take into account a lot of important factors.
Who cares how much it costs Amazon to make the thing? You don't like the price, you don't buy it. What do Amazon's costs have to do with that decision?
I'm not sure why you phrase it so discouragingly. What's wrong with wanting to know simply for curiosity's sake? I'm interested in buying a kindle. I'd be interested in buying it without knowing individual component costs. But if offered the chance to see the prices, I'd be interested, sure.
My phrasing was because of the markup discussion above. I'd assumed people were talking about the high (?) markup because it'd factor into their buying decision, which seems pointless.
I see nothing wrong with wanting to know for curiosity's sake.
As a consumer, I'm more interested in cost of ownership of the device in comparison with just buying physical books rather than the underlying manufacturing cost.
The research guys, iSuppli, rightly did not include them - they're usually paid by the chip vendor to ARM (or whatever licensor), and included in the chip price. The writer tried to be clever, and missed.
Retail price = 3 x BOM (bill-of-materials) is one rule-of-thumb for volume hardware without special assembly or extraordinary software.
Well, iSuppli did not mention them. The writer brought it up: "Royalty Payments Not Addressed ... One cost iSuppli's teardowns don't address: Any royalties paid to ARM."
He missed the mark - he might have had a point for data formats. I think that licensing costs are borne by the manufacturers (of the final product) for firmware they toss in to decode DVDs, MP3s, etc. Cf. recent TomTom/Microsoft FAT kerfuffle.