IEEE Solid-States Circuits Magazine - Summer 2023 - 84

As the industry continues to scale, it must
adopt more sustainable practices to minimize
its impact on the environment and ensure
long-term viability.
To address these challenges, a true
system-architecture-to-technology optimization
loop is required. An STCO
framework can address the right technology
ingredients that can unlock
the major system scaling bottlenecks
and enable new ways of computing.
In addition, STCO can enrich the technology
road map with system scaling
challenges and application-specific
requirements (Figure 3).
To achieve optimal system implementation,
multiple technologies,
such as extreme CMOS logic scaling,
advanced 3D packaging, novel memory
elements, and even Si photonics,
must be optimized in a subtle manner.
Previously, these different technology
research activities have been
conducted separately, each with their
own DTCO research and road map.
Hardware/software codesign will become
ubiquitous.
In the next 75 years, STCO will
play a critical role in the semiconductor
industry's evolution. Solutions
will be driven by application-specific
requirements, and optimal system implementation
will require a subtle optimization
across multiple technologies.
The move toward STCO represents
a significant shift in the way that engineers
and researchers need to approach
the semiconductor industry.
Traditionally, engineers were trained
in one particular area of expertise,
such as materials science, electrical
engineering, or computer science.
However, to truly optimize systems
and technologies for future applications,
engineers and researchers will
need to develop a more multidisciplinary
skill set that spans multiple
areas of expertise.
This means that the next generation
of engineers and researchers
will need to be comfortable working
across different disciplines and collaborating
with colleagues from a variety
of backgrounds. They will need
to be skilled in areas such as materials
science, electrical engineering,
computer science, even biology, and
many more.
To prepare for this future, universities
and other educational institutions
will need to rethink their
engineering curricula to ensure that
they are providing students with the
necessary interdisciplinary skills and
knowledge. This may involve developing
new programs or courses that focus
on STCO or incorporating STCO principles
into existing courses. Ultimately,
the success of STCO will depend on the
ability of engineers and researchers to
work together across disciplines and
leverage their combined expertise to
develop optimized systems and technologies
for the future.
Beyond von Neumann
It is hard to imagine the future of
IC technologies without considering
the limitations of the von Neumann
architecture, which consists of a
processing unit, memory, and I/O
devices. The processing unit contains
the CPU, which executes instructions
and performs arithmetic
and logic operations. Memory stores
both instructions and data that the
CPU uses for processing, and I/O devices
allow the computer to communicate
with the outside world.
While this architecture has been
the foundation of computing for decades,
it is facing challenges in the
era of big data and AI. The traditional
von Neumann architecture, where
the CPU and memory are separate,
creates a bottleneck in the processing
System Optimization Based on Off-the-Shelve Technology
Technology
Components at
Device Level
Impact System
Performance
STCO
Specific Technology
Developments that Intercept
System Challenges
Technology Optimization for Generic Applications
FIGURE 3: The STCO framework.
84 SUMMER 2023
IEEE SOLID-STATE CIRCUITS MAGAZINE
New System Architectures
Built on New Technology
Capabilities
System-Level
Architecture and
Applications
Drive Technology
Roadmap
STCO
Technology Scaling
System Scaling
IEEE Solid-States Circuits Magazine - Summer 2023

Table of Contents for the Digital Edition of IEEE Solid-States Circuits Magazine - Summer 2023
