OUM / fuel cell cover stories in the venerable e.e. times of 16 july
in the 16 july paper edition only, columnists lammers and wade
run down the intel flash alternative scene. pertinent points:
- on polymer memory (PFRAM), intel's lai estimates
chip costs at "about one-eighth the cost of CMOS memory"
or 4-to-8 times less for a flash memory card replacement.
- for polymer, write time is 50 usec, vs. 1 ns for NAND flash.
- ovonyx alone has 20 full-time engineers for OUM.
- lowrey & lai both estimate 3-5 years until commercial production.
- regarding flash scaling, experts at the june VLSI
tech symposium in kyoto see scaling difficulties beyond
70-nm that goes to manufacturing in 2005, plus a definite
need to find a new floating-gate insulator by then.
- curiously, "all of the flash memory papers dealt with
the challenges of embedding flash with logic, rather
extending discrete flash devices."
as an aside, a market research source mentioned in passing
the 1970 gordon moore amorphous memory paper in conjunction
with the perception that ovshinsky wasn't taken seriously
then due to his "grandiose" memory cell claims.
_____
bonus article: ethanol (non-PEM) portable fuel cells improvement
report by medis technologies ltd., claiming 150 -> 450 W-hr/kg.
contains favorable commentary by competitor bob hockaday,
chief scientist for manhattan scientifics inc. that the medis
electrolyte boost, if proven efficient, could scale to autos
and power plants. the eyebrow-raiser for the author was
that the company was actually proposing OEM cell prices of $15,
given manufacturing costs at $9 (size of cell not stated, but
supposedly yielding 20-hours cell phone talk time). |
