Computer Architecture for Embedded Systems (CAES)

Faculty of Electrical Engineering, Mathematics and Computer Science (EEMCS)

Section Head: Prof. dr.ir. G.J.M. Smit

Keywords: Reconfigurable, composable, energy-efficiency, mapping tools, dependability, streaming applications, optimization

The main emphasis of the group is on efficient architectures for dependable embedded systems. Within this theme the chair performs research on three related key areas:

1)	efficient multi-core architectures for adaptive streaming applications,

2)	using ICT systems for efficiency, in particular optimization algorithms for energy management in buildings

3)	dependability issues of embedded systems.

Energy-efficiency is the main drivers of our research. Energy-efficiency is important for streaming applications found in battery powered consumer devices (e.g. i-phones or portable multimedia players), for high-performance professional equipment, and is of crucial importance for wireless sensor networks.

ad 1)

The effort on energy efficient architectures focuses on reconfigurable processors for streaming applications, e.g. found in consumer products such as media players. However, within high performance embedded computing (medical imaging, radar processing), extrapolating the current trend of using general purpose processors for future systems predicts excessive power consumption and dramatically reduced reliability. MPSoC (Multi Processor Systems-on-Chip) devices for streaming applications are prime candidates for use in this application domain as well.

ad 2)

The focus of using ICT for efficiency is to incorporate ICT technology into buildings to improve the overall energy efficiency. By designing so called smart grids the overall energy efficiency can be significantly improved by choosing the best energy generator/source available at the moment. In this research we use a combination of prediction, planning and real-time control.

In MPSoC (Multi Processor Systems-on-Chip) systems for efficient streaming applications dependability techniques play an important role. One of the problems in manufacturing a MPSoC with a billion transistors using deep-submicron technologies (65 nm and below), is an increase in the probability of defects which results in a decreased manufacturing yield and lower reliability. To effectively deal with this increased defect density, we need efficient methods for fault detection, localization, and fault tolerant architectures implemented on-chip.

Key persons: prof. dr. ir. G. Smit, Prof. dr. ir. Marco Bekooij (0.2 FTE), dr. ir. H.G. Kerkhoff, dr. ir. S. Gerez, dr. ir. A. Kokkeler, dr. ir. J. Kuper

Projects:

European funding:

·	CRISP: Cutting-edge Reconfigurable ICs for Stream Processing (EU/FP7)

·	FP7-So(o)s

·	FP7-ADVANCE

·	TOETS (Catrene)

National funding:

NWO-Easy

STW-Nest

STW-SFEER

·	STW-CMOS beamforming

·	STW-SeaSTAR

·	STW-Ad rem

FES-STARS

·	Pieken in de Delta-DTFC

·	GO-EFRO-TSP

·	DREAM: Dynamic Real-Time control of Energy streams in buildings (NWO/STW Smart Energy Systems)

Industry:

·	ESA (1 PhD)

·	NXP (3 PhD)

·	Thales (1 PhD)

·	E.ON UK (1 PhD)

CAES publications