I'm sure you're correct bringing up yield on large chips but you're
overstating the case in pointing to a 1/4 size chip. Winsor's
comparison was to the D300 which has a DX size chip. Still bigger than
4/3 (and possibly even very significant from a cost standpoint) but not
nearly as expensive as a full frame sensor.
Chuck Norcutt
Moose wrote:
> Winsor Crosby wrote:
>
>> The 4/3 sensor could not be that much cheaper to produce.
> I'm out of date. I know that it used to be true that yield, and thus
> price, on a chip 1/4 the size made a difference of much more than four
> times in unit cost. It's simple mathematics on the rate of flaws per
> area of the wafer vs. individual chip size. One flaw ruins 25% of four
> smaller chips in a given area but 100% of one big chip. So, depending on
> the relationship of flaw incidence and the size of the larger chip, it's
> even possible to have a yield rate of about 0% for the large chip and
> 50% or better for the smaller one. Not saying that's what's happening
> here, but it was a reason for very high initial prices on some
> microprocessors in the past.
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|