Profiling and benchmarking of embedded Java Virtual Machines.
Dynamic power management of digital circuits and systems.
CAD for low-power integrated circuits.
Signal integrity in deep sub micron integrated circuits.
Signal integrity is a critical issue in deep sub micron integrated circuits,
because of reduced noise margins and increased noise sources. Accurate
characterization of noise sources and effects is needed to support
integrated circuits design. On the other hand, the behavior of a noisy signal cannot be captured from outside a chip. Our research activity is aimed at enabling accurate on-chip characterization
of noise effects and wiring parameters. This is done by realizing test chips that integrate both the structures under test and the measurement equipements,
providing a digital external interface for measure control and data acquisition.
Cooperations: ST Microelectronics - DI, University of Ferrara
Computer simulation of biological neural networks.
Intellectual-property protection in digital system design.
Sub-micrometric integration technologies provide the opportunity of realizing
micro-fabricated devices that interact with biological molecules. BioMEMs
are hybrid micro-fabricated systems that can integrate on the same chip
components of a different nature: electronic
(electrodes, transistors, ...), mechanical (cantilevers, switches, ...),
fluidic (channels, pumps, ...), chemical (probes, biosensors, ...), optical
(mirrors, ...). In principle, BioMEMs enable the realization of what is commonly called a lab-on-a-chip, integrating all the instrumentation required to perform biological analysis.
Our research activity in this field is aimed at the realization of
microfabricated biosensors for studying the interaction between bio-molecules
(DNA and proteins).
Cooperations: DEIS, University of Bologna - STI, University of Urbino - Department of Physics, University of Ferrara
Enabling technologies for ubiquitous geographic information systems (GIS)
Advanced algorithms for string comparison and DNA sequencing
Parallel implementation of string-comparison algorithms
String comparison is a critical task in many fields: DNA analysis, speech recognition, associative memories, ... The similarity between two strings is usually evaluated in terms of edit distance, that is the cost of the edit steps (insertion, substitutions, deletion) required to make the two strings equal to each other. Edit distance can be evaluated by means of dynamic programming with complexity O(M*N), where M and N are the lengths of the two strings. Parallel implementation can achieve a performance that is O(M+N).
Our research activity in this field is aimed at the parallel implementation of dynamic programming algorithms on reconfigurable hardware platforms
(FPGAs, systolic arrays, ...).
Cooperations: EPFL, Lausanne - STI, University of Urbino - DI, University of Ferrara
|Last update: 9-2002
||alessandro.bogliolo @ uniurb.it