IEDM keynotes: Si's future, power's potential, bioelectronics breakthroughs - Solid State Technology
This article is from

IEDM keynotes: Si's future, power's potential, bioelectronics breakthroughs

by Michael A. Fury, Techcet Group

December 8, 2010 - The 2010 IEEE International Electron Devices Meeting (IEDM) opened on Monday, December 06 at the San Francisco Hilton with a record attendance of 1645. The makeup of the submitted paper pool is shifting, with industry submissions down by 50%, university submissions up, and government submissions holding steady. This likely correlates to the industry consolidation that has taken place over the past decade, with leading edge R&D being conducted by far fewer companies now, and with greater reliance on university and consortium collaborations. However, the percentage of papers selected for presentation in the meeting remains equivalent at 48% for both industry and academia.

The president of Samsung's SAIT, Kinam Kim, gave the first of three keynote presentations, his on the challenges and opportunities for future silicon technology. He believes that DRAM will be extended to 10nm, and that 20nm NAND flash is ready for volume production now. 3D NAND is extendible to <10nm with existing technology by 2013-2015. Given the increasing aspect ratio of gate stacks, supercritical CO2 is once again evoked as a critical enabling step. Beyond NAND flash, he believes ReRAM will prove to be the best choice for both current NAND applications and for universal memory. We will have off-chip optical interconnects by 2015.

Energy efficiency enabled by power electronics was the keynote topic of Arunjai Mittal of Infineon. Windmills intrinsically generate AC current, which can be transported well enough by conventional separated AC transmission lines on land. Wind farms at sea require coaxial transmission lines to land, which generate 30% transmission losses with AC. Large rectifier systems have been developed to consolidate the wind farm power, convert it to DC for transmission to land, and reconvert it to AC for the grid, with a net loss of 15.5%. Braking systems on trains in the EU are designed to generate electricity and recapture some of that energy. Taken together, EU train braking generates 30GW/year. Lighting accounts for 20% of global energy consumption, and fluorescents are still the best value in terms of lumens per watt. New digital circuitry is being implemented in fluorescents to adjust their power usage to compensate for chemical degradation effects as the bulbs age. The electricity consumption of IT data centers exceeds the total usage of many small countries, such as Belgium and Holland. A typical data center power bill is in the range of $1.5M per month. Per month! Most of this is for air conditioning, and Infineon and others are looking at ways to reduce this energy consumption as well.

In a significant break with the past, the final keynote was delivered by Luke Lee of UC Berkeley, speaking on bio-nano science and technology for innovative medicine. A keynote on the crossover of electronics and biotechnology signifies a certain coming of age for this field, which had been growing slowly but quietly for the past few decades. Luke spoke of satellite nanoscopes, invoking the telescopes of Galileo which opened the heavens for exploration (although he originally invented the telescope to improve commerce in Venice by identifying at a distance the ships approaching port; stargazing came later). The current research has produced nanocrescent optical antenna devices, which are hollow gold spheres with an open end that can be tuned to resonate when different frequencies of light are shown on them, thus service as biological markers when they are chemically treated and attached to specific sites. In other work, he has developed a silicone microfluidic device capable of capturing and individual live cell and measuring its ionic current response to various chemicals and pharmaceuticals. This opens the door to more comprehensive testing on fully functional cells.

In my next entry, I will begin my comments on the submitted presentations and other aspects of the meeting.

On Sunday, December 05 before the technical conference officially got underway, a full day of short courses attracted 580 attendees to two separate tracks. Guido Groesenken of IMEC organized one track under the topical heading of "Reliability and yield of advanced integrated technologies."  The five presentation units were:

  • FEOL Reliability: BTI and TDDB in high-k/metal gate, FinFET and Ge-based technologies; Ben Kaczer, IMEC.
  • BEOL Reliability: metal migration, stress voiding, via and low-k dielectric reliability; Shinichi Ogawa, AIST.
  • Electrostatic discharge aspects of FinFETs and other most advanced CMOS technologies; Christian Russ, Infineon.
  • Reliability-aware design of integrated circuits; Muhammad Alam, Purdue.
  • Yield, yield models and design for manufacturability; Andrzej Strojwas, Carnegie Mellon & PDF Solutions.

The other track, organized by Kelin Kuhn of Intel, focused on "15nm CMOS Technology," and also comprised five units:

  • CMOS technologies -- trends, scaling and issues; Thomas Skotnicki, ST Microelectronics.
  • 15nm device challenges and solutions; Mukesh Khare, IBM Research.
  • Lithography for 15nm technology node; Sam Sivakumar, Intel.
  • BEOL technology toward the 15nm technology node; Yoshihiro Hayashi, Renesas.
  • Device / circuit interactions at 15nm technology node; Clive Bittlestone, TI.


Michael A. Fury, Ph.D, is senior technology analyst at Techcet Group, LLC, P.O. Box 29, Del Mar, CA 92014; e-mail [email protected].


Recent Semiconductors News

Solid State Technology Article Categories:

Semiconductors Materials
Industry News Subsystems
Wafer Processing Facilities
Lithography Wire News
InspectionSST Magazine Current Issue
Device ArchitectureSST Archives