 |
TECHNOLOGY
Semiconductor technology lies at the heart of
the amazing revolution we are witnessing in computing, communications,
consumer electronics, transportation and health care. This revolution
is enabled by designing and building successive generations of chips that
perform an ever increasing number of functions, run faster, and cost less.
In the course of this progress, the semiconductor industry entered the
nanotechnology era in the year 2000, and by 2004, the industry was shipping
devices with physical gate lengths less than 40 nm and insulator thickness
less than 1 nm. This reinforces the industry's position as a true nanotechnology
pioneer, through continued technology advances at the pace of Moore's
Law. This progress will enable future functionality and applications that
we perhaps cannot even imagine today.
To sustain this remarkable pace
of technology progress, we must be able to keep pushing the technological
limits of semiconductor design and manufacturing even further. To build
tomorrow's systems, we must learn to innovate swiftly with new design
methodologies, devices and materials. This requires a large and coordinated
effort in research and development among corporations, governments, and
universities.
To this end, the SIA forms consortia and other partnerships to fund advanced
research and pool resources and ideas. The goal is not just to maintain
the technological leadership of U.S. semiconductor companies, but also
to keep the innovation engine of the country and the economy vibrant and
strong. The SIA pursued four major technology initiatives this past year:
Continuing our involvement with the Focus Center
Research Program - a University-based research program funded jointly
with the Department of Defense, focused on cutting-edge research to take
CMOS technology to its ultimate limits.
Launching a new Nanoelectronics Research Initiative,
in partnership with the National Science Foundation, focused on "beyond
CMOS" nanotechnology.
Publishing the 2004 update to the International Technology
Roadmap for Semiconductors.
Lobbying and building partnerships to increase federal
spending on education and research in the physical sciences, mathematics,
and engineering.
The International Technology
Roadmap for Semiconductors
Working with the Semiconductor Research Corporation (SRC) and International
SEMATECH, SIA helps semiconductor companies collaborate on technology
challenges. One way we do this is by publishing the International Technology
Roadmap for Semiconductors (ITRS), which identifies industry trends, highlights
technical obstacles, and helps companies align product cycles with developing
technologies.
The 2004 ITRS update represents the 7th international version of this
roadmap, reflecting input from nearly 1000 experts and researchers from
Europe, Japan, Korea, Taiwan, and the USA.
Innovation Highlights
2005 brought exciting developments in semiconductor technologies:
- Despite significant technology challenges, the industry continued to maintain the pace predicted by Moore's Law – the doubling of transistors every two years.
- Transistor speed continued to improve at the historical improvement rate of 17 percent per year, although the challenges became more complex due to a concurrent increase in leakage currents.
- Mobility enhancement techniques will allow transistor performance to improve as required without the introduction of high-k gates and metal electrodes materials for 65nm and maybe for 45nm.
- Lithography likely will not be a limiter down to 45nm. 193 nm, the current technology, will be challenged only at the at 32/22nm node by Extreme Ultra-Violet (EUV) lithography. Exposure tools with 193nm wavelength will dominate the next technology generations (65nm, 45nm and maybe 32/22nm) with both dry and immersion solutions. Control of critical dimension (CD) below 4nm will be a major issue.
- Partly due to lithography, and partly due to the inherent variability at the tiny dimensions of state-of-the-art devices, design-for-manufacturing (DFM) is growing rapidly in importance.
- Research intensified on major technology innovations like high-k dielectrics, metal gate electrodes, and multiple-gate MOS transistors, which are forecast to enter manufacturing within a decade. These represent major shifts, where some basic device materials and structures will undergo change for the first time in more than 30 years.
- The low-k interconnect dielectric roadmap is essentially unchanged for the first time in 10 years. Materials with k~2.7 are in manufacturing and k~2.4 materials will follow.
- Multiple new package solutions will be required to increase functionality and to reduce cost will require additional attention in the future.
- Mixed-signal and analog chips continue to grow in importance, driven especially by consumer and communications-related markets.
- Discussion has been initiated for possible introduction of 450mm wafers in the 2012 timeframe.
- There is a growing focus on emerging research devices and materials, as reflected in the newest roadmap chapter.
Basic Knowledge Fuels Industry
Progress
To overcome formidable obstacles in design and manufacturing of devices
and circuits, the semiconductor industry must conduct advanced research
and train graduate students in all technology disciplines. That's why
SIA formed the Semiconductor Research Corporation which, since its inception
in 1982, has graduated more than 3,000 advanced degree students.
Focus Center Research Program
In collaboration with the US Government, SIA created the Focus Center
Research Program (FCRP) in 1999. This program brings together the US semiconductor
industry, the federal government (Department of Defense), and 30 of the
nation's most prestigious universities. They collaborate on cutting-edge
research deemed critical to the growth of US technology industries. It
is the most ambitious research project the US chip industry has undertaken
since SIA companies formed the SEMATECH consortium in 1987.
FCRP researchers investigate technology solutions for key issues (as highlighted
by the ITRS) that will arise eight to ten years in the future. The FCRP's
five national focus centers channeled over $25 million in 2003 into new
research activities in these areas and expect to spend some $37 million
in 2005. Importantly, representatives from the organizations that fund
each center work with research teams to bring these advanced new technologies
to market.
Here's a flavor of the exciting research currently taking place at these
focus centers:
- System Design
- The Gigascale System Research Center created and developed Platform-Based
Design as a paradigm-shifting, design methodology for complex systems-on-a-chip.
GSRC also began work on a living systems roadmap through close collaboration
within GSRC, industrial partners, and the SIA roadmap effort.
- Interconnects
- The Interconnect Focus Center (IFC) developed an interconnect framework
for 40Terabit/s optical bandwidth, a 3D integration process for hyper-integration
of devices in a circuit, and continued work on optical interconnects
and on innovative micro-channel cooling for "hot" chips.
- Circuits
- The Center for Circuits and Systems Solutions (C2S2) addressed the
formidable power and leakage issues using novel circuit techniques,
and demonstrated 33 percent energy saving and 43 percent leakage reduction
for an example technology. C2S2 also developed the via patterned gate
array (VPGA) as a new solution to address the manufacturing complexity
and process variability issues for application specific interconnects.
- Devices -
The Center for Materials, Structures and Devices (MSD) developed and
demonstrated significantly improved device mobility (speed) with a strained
Si channel, novel high-K dielectrics with a Ge channel MOSFETs, and
the first integration of nanotubes with CMOS technology.
- Nanomaterials and nanodevices
- A fifth focus center, Functional Engineered Nano Architectonics (FENA),
commenced operation in 2004 with an emphasis on nanomaterials and nanodevice
research.
These centers are now working with an integrated
research agenda to ensure efficiency and reduce redundancy in the research
effort for scaling CMOS technology to its ultimate limits.
Nanoelectronics Research Initiative
Most experts agree that CMOS will reach the end of its progression in
about 15 years - hitting physical, technological and economic limits.
The Nanoelectronics Research Initiative (NRI) is proposed as a mission-oriented
platform to accelerate and augment research "beyond CMOS" technologies.
The Technology Strategy Committee of the SIA has defined its mission as
follows:
"By 2020 discover and reduce to practice via
technology transfer to industry novel non-CMOS devices, technology and
new manufacturing paradigms, which will extend the historical cost/function
reduction, along with increased performance and density for another
several orders of magnitude beyond the limits of CMOS."
The effort would involve the government, industry
and academia working together to link existing efforts, to identify gaps
and to seed research to bridge the gaps; and to demonstrate proof-of-concept
for a few, select ideas.
Reversing the Decline in Federal Funding
In the 1990s, federal funding declined precipitously in the areas most
critical to our industry's continued success: the physical sciences, mathematics,
and engineering. This has seriously reduced the number of faculty and
students in these disciplines, slowing the pace of university research
and creating a shortage of skilled workers for our companies. In the past
year, SIA has energetically addressed this problem:
- SIA has engaged deeply with government agencies
to ensure that sufficient research resources are committed to maintain
progress along the path projected by Moore's Law.
- SIA built a partnership between the NSF and SRC
that will support significant funding of ITRS-related research.
- SIA was successful in increasing government funding
of the Focus Center Research Program to $17 million in 2004.
Innovation Demands Investment
We rely on semiconductor technology and take it for granted in our everyday
lives. But the marvels of today are really the fruits of research seeds
planted decades ago-investments that either funded discoveries and new
technologies or helped educate the very engineers and scientists who now
form our workforce. The very fact that these advances required decades
of investment stands as a warning against complacency in our future investment
strategy.
Our future performance as an industry depends on our capability to create
new knowledge and develop it into technologies that drive our economy,
guarantee our national security, and improve health and the quality of
life. Only with significant government support of R&D can we progress
on the current course of miniaturization and address nanotechnology's
enormous challenges with tremendous benefit for all.
|
|
|
