Our Work

Our recent focus has been on the design and analysis of non-traditional computing systems (including optoelectronic, parallel, and reconfigurable architectures); analog and digital VLSI design (in CMOS, SiGe, SoS and SoI technologies); and computer aided design methodologies (multi-domain mixed-signal simulation, high-level synthesis, and formal and temporal verification). Previous funding has been in the areas of multi-domain, mixed signal modeling and simulation environments, high speed electrical interconnect, and photonic circuits implemented in SiGe. Recent projects include adaptive signal processing for radar, multi-frequency dielectrophoresis for manipulating biological particles, and oscillator based computing.



Copies of Papers



Present Research

Oscillator-Based Computing

The Oscillator Technology Collaboration is a venture with three separate groups: Visual Cortex on Silicon [NSF], Sensing and Computing with Oscillating Chemical Reactions [NSF], and Ultra Low Power Non-Boolean Systems [Intel].





Previous Research

Chatoyant

Chatoyant is a multi-level simulation environment that supports the design and analysis of multi-domain micro-systems.  These include systems that span electronics and mechanics (MEMS) as well as opto-electronics (OMEMS).

High-speed Interconnects

Differential signaling techniques provide high speed at the expense of "two wires per bit" for information transfer.  Multi-bit differential signaling amortizes the cost of a reference value across all the information bits by the use of balanced codes.

Opto-Electronic Computer Architecture and Design

The fundamental advantages of optics as a communications medium for information processing is not in the "speed of light," but rather the unique qualities of light in its non-interference, low temporal dispersion and information capacity.  The use of both free-space and guided wave optics provide many opportunities for improving the connectivity and functionality of future computing systems.

VLSI Computer Aided Design

Design Automation is essential for the development of modern digital systems.  Computer aided design encompasses the design capture, simulation, synthesis, analysis, and testing of complex systems.

Learning and Prediction

The phenomena of spatial and temporal locality in the behavior of executing programs enable the prediction of both data access patterns and data values.  In a similar way, communications patterns in communicating parallel computers can be predicted and hardware resources can be optimized to minimize communication latency.

Computer Vision Architectures and Algorithms

Some of our earlier work on VLSI based architectures for computer vision was based on the use of content addressable parallel computers for associative processing.






Other Research