Quote from: runeks on September 16, 2013, 05:34:46 PM

Quote from: Vycid on September 16, 2013, 05:30:40 PM

Quote from: runeks on September 16, 2013, 03:51:36 PM

Quote from: hammurabi on September 16, 2013, 03:36:32 PM

Quote from: velacreations on September 16, 2013, 01:44:40 PM

Quote from: runeks on September 16, 2013, 12:31:10 PM

friedcat has stated that "final margin cost for 1TH/s is less than 10k$":

https://bitcointalk.org/index.php?topic=99497.msg2013118#msg2013118

I wonder if that price is still accurate. He made that statement in May, and since then, he's supposedly purchased like 200TH or more worth of chips (or maybe more)

Those are rather initial costs since volume and mask process changed since then.

What do you mean by "mask process"?

I think he is referring to the NRE for the Gen 1 mask set and is ignoring the Gen 2 NRE on masks.

friedcat is talking about marginal cost, so I don't see why NRE should be included in that figure. But it's entirely possible that the price has changed since then, I'm no expert on what takes up the marginal cost of producing a chip.

Well, if you want to ramp up production with more than one manufacturing line, you might need a second mask set. For FC's volume that should be totally unnecessary, even for Gen 2.

Marginal costs shouldn't change much for the same process unless you make a new deal with the foundry.

Anyway, after actually MAKING the hardware (and doubtless encountering some speedbumps), I do wonder how accurate that estimate is.