Reitarius, here is some more discussion on OUM.
You noted:
<<i can see why t. lowrey wants to go for embedded system RAM first,
because it might just be a similar 25% process cost increment for an
extra 3-5 process steps as IBM claims for combined embedded (volatile) DRAM and logic. if ECD hits the same stride and can also offer
nonvolatility here, then they can save chip area from having
embedded FLASH, *as long* as that all-to-critical cell area is submicron. >>
There is also another advantage - replacing those ssllooooww hard disks. Here is a quote from an article by Mark Johnson of US Naval Research Labs:
"PC users may start to benefit from this startling progress early in
the next century. "Spintronic devices would allow us to replace the
redundancy and cost of having two memory systems--RAM
[random access memory] and hard discs--by a single, fast spintronic
device that does the work of both," says Johnson. Accessing data
on a hard disc is a time-consuming business. This is why computers
take so long to boot up. With a spintronic memory, they would be
ready to work almost immediately. "Making these devices is pretty
simple, and people are talking about making 1 gigabyte of
nonvolatile RAM for the same cost as today's 1-gigabyte hard
drives," says Johnson. That is, for well under $200. "
The full article by Johnson is copied into my post #3518 (which I forgot to mention to you). ECD also claims hard drive replacement as a goal. Why not? They apparent;ly have a fundamentally better approach.
As for RAM, I think the OUM might well be cheaper than the present chips due to not needing the capacitors used in DRAM.( And, apparently, an OUM cell can store 4 bits of information.) My comments from 3518:
<The Spintronic switch (i.e., a memory cell) is touted (see article below) as "could hardly be simpler". However, the Ovonic switch is simpler yet! Not counting the substrate layer and any protecting overcoat layers, here is what I think each device requires in the way of layers of materials. (Essentially, each feature of a chip device is formed from a layer that is then pattern-etched). The spintronic device has inherently five or six layers - three for the switch, two for the read wires (which are on opposite sides of the switch surfaces), and one or two for the write wire(s). The Ovonic switch has inherently only two layers - one layer for the switch and only one layer for the two wires (because they are on the same side of the switch layer) that serve both read and write functions. Tyler is emphasizing the EXTREME simplicity of the OUM, and we clearly win this important point.>
Since capacitors require etching deep pockets on the chip, to get enough charge storage area in a small region of the memory element, OUM chip processing might be less costly. I think maybe the additional processing steps Lowery has mentioned are only for making integrated logic/memory chips - by adding the Ovonic resistive spots and leads to logic chips. It is not clear to me that more-than-usual processing steps are needed for OUMs. |
