Re: Direct Compact Flash to Hard Drive




"Ron Hunter" <rphunter@xxxxxxxxxxx> wrote in message > Yes, indeed. The
price for 1GB of SD ram has fallen from just under $20
to under $8 in less than a year. It looks like flash ram will go to a few
cents/GB in a couple of years. Nice trend. Now if only some other
things, like digital picture frames, would see a similar drop in price. It
is hard to understand how electronic pricing works. Why can a device like
an iPod sell for under $200 with Gigabytes of ram, and a small display,
while a basically similar device to store digital images costs 5 times as
much? Still, HD storage is in the $.30/Gigabyte range, while flash ram is
still over 20 times as much. Seems like the solid state device would be
less expensive to manufacture.

(Caution much of the following is opinion and speculation so take it as a
point of conjecture to begin from. Not a hard and fast for-sure answer.)

I suspect that part of the higher price for anything that includes a display
is the cost of waste. If a solid state memory has one or two bad bits the
software can easily just ignore those sites and use the ones around them.
But if a single display element is bad in an array you will have a possibly
unacceptable blank pixel.

The larger the memory the more bad bits there are likley to be but the
easier it is to work around them. But the larger the display the higher the
odds that one or more bad pixels will be noticable. So if one objectionably
bad display out of 10 made is happening, the cost of making that unsellable
display will then be portioned out to the other 9. And at one time I heard
that the proportion was actually closer to 50% bad. Now this was some time
ago that I heard this and so the rejects are probably fewer now, but it is
impossible to make such rejects totally go away. And if the cost to make a
single large display is the same as making a single wafer of a dozen smaller
displays, and if 1 in 10 of the small displays is rejected, the entire wafer
cost is not wasted. But a single large display on a single wafer will incur
the entire cost of manufacturing and thus rejects will raise the cost of all
the similar displays.

Since memory can adjust for bad spots but displays can't it is probable that
memory costs will continue to fall much faster than large display costs.
Maybe if they figure how to make a single large display from a large number
of small displays, seamlessly, some of the large costs of larger displays
can drop at a similar rate to memory.

JMHO

Randy


.


Quantcast