This, now old, "news" concerning the development of Ferroelectric FLASH in Japan doesn't seem to want to die:
engadget.com
In trying to figure out why the stories keep coming, I found this reference to a Sandisk deal with Toshiba:
electronicsweekly.com
Interesting because the specs suggested by AIST's FeFET FLASH would surely put a crip in Sandisk's "cash cow" NAND products and licensing. Even more interesting because, despite the Sandisk article's gross speculation concerning the "3D" memory in question being the old Matrix one time programmable Flash IP, it seems obvious to me that no OTP NV memory is not going to cover Sandisk's rear end, products, or massive royalties currently tied to the standard mainline NAND market.
So if the "3D" technology in the Sandisk/Toshiba deal is in fact worth the time of day... the "3D" in question may actually be a reference to a Toshiba spin of the AIST FeFET NAND.
Its certainly not... a certainty... but Toshiba is a member of AIST and must have been aware of the Tokyo Univ work for some time. Moreover they know of Spansion's process advances with their NOR-NAND combo IP, and must be almost as concerned as Sandisk is about defending their NAND markets from its encroachment. In any case all the above speculation and ink being devoted to the FeFET NAND combo development may well be misdirected.
This "old news" from the same date is of far greater import to the future post FLASH world of NV memory... IMVOP:
physorg.com
This quantum dot IP the Koreans and Germans have come up with is a truly remarkable break through. They've used a perforated aluminum oxide mask to "enable" the regular self-alignment of PZT quantum dots without the normal lithographic tools. This means probe storage densities without the need for a MEM device probe.
This would be at least comparable if not greater densities than could be achieved by the memristor, capacitor-less IP that HP is shooting for or the Ferroelectric Polymer IP Intel's been researching for eons. No doubt it would take longer to develop than a FeFET NAND, but you're talking massive price, performance, and power savings for any almost application.
Who knows how many other ways there may be to persuade PZT cells to "self-align in regular patterns for practical circuits... remember graphene? <g>
I any case... I'm betting we get news of a "new" Ferroelectric product from Samsung or Toshiba... maybe both by next March. I doubt that it will be either of the above technologies because it would do too much damage to their FLASH sales.
But if you see someone start producing commercial quantities of standalone F-RAM "embedded" as part of SSD chip-sets... FeFET NANDs will be right around the corner.
And by 2015... assuming you're still alive... and give a Hoo Hoot... just remember you heard about the "quantum dot" F-RAM here first. <vbg>
0|0 |
