From: EETimes

8. Resistive RAM or the memristor

The pursuit of the universal memory goes on. It needs to be simple like a DRAM, or preferably even simpler as those capacitors are a problem to scaling. It needs to be able to retain data for years with the power off and able to be used millions of times. It needs to be simple to make using conventional methods and with materials that are not out of place in conventional wafer fabs. And we still haven't found it yet.

Or have we?

In 2009 Unity Semiconductor Corp. emerged from seven years of stealthy research with its conductive metal oxide (CMOx) technology, although we are pleased to note that EE Times was reporting on Unity in April 2006. But 2009 has also seen the arrival on our radar screens of 4DS Inc., Qs Semiconductor Corp. and Adesto Technologies Inc.

We are also aware that many of the larger IDMs are active in RRAM. And the reference to the memristor is because two-terminal devices that display a memory-effect in their resistance characteristic are effectively the practical implementation of the theoretical work, championed by Hewlett-Packard Labs, on the memristor, often described as the fourth passive circuit element after resistors, capacitors and inductors.

9. The through-silicon via

The depth of the interconnect stack on top of the leading-edge silicon surface is deep and can vary markedly in minimum geometry. We have speculated that this could result in a splitting of front-end fab production into surface and local interconnect followed by higher stack connection, possibly in different wafer fabs.

The desire, for marketing as well as technical reasons, to mount multiple die in single packages is also driving a need for more sophisticated interconnect and the arrival of the through-silicon-via passing completely through a silicon wafer or die is clearly important in creating 3-D packages.

In May 2009 Austriamicrosystems started producing TSV parts on a foundry basis, targeting suppliers of devices for 3-D integration of CMOS ICs and sensor components. Expect more of the same in 2010.

10. Various battery technologies

We have become so used to Moore's law and the steady miniaturization of microelectronics it is easy to become frustrated with a technology that does not double in performance every two years. But battery technology is relatively mature and is not driven by the same forces as the integrated circuit. Indeed if energy storage becomes too dense it can become dangerous.

Nonetheless we all rely increasingly on batteries for energy storage and to power our various gadgets. Indeed it is arguable that without further breakthroughs in battery technology for electric vehicles the compatibility of the automobile and sustainable green transportation is in jeopardy. So the pressure is on.

Recent spins on nickel- and lithium-based battery chemistries, such as nickel oxyhydroxide, olivine-type lithium iron phosphate and nanowires, are gunning to displace the venerable but problematic alkaline-manganese dioxide formulations. ReVolt Technology, a developer of rechargeable zinc-air batteries, has selected Portland, Oregon as the location for its U.S. headquarters and manufacturing center. We expect similar developments to come on apace in 2010, and every smart battery is set to provide a power management IC opportunity.