A5032461807- Embedded Systems Design Techniques
- Parallel Computing and Optimization Techniques
- Interconnection Networks and Systems
- Real-Time Systems Scheduling
- Meteorological Phenomena and Simulations
- Model Reduction and Neural Networks
- Numerical methods for differential equations
- Formal Methods in Verification
Lund University
2022-2024
We introduce a novel approach for energy-efficient scheduling of data-dependent stream programs with packet types on multicore architectures voltage and frequency scaling. To have the given application meet specific throughput demands while minimizing energy consumption, we enhance existing crown by packet-type dependent parameters. Formulation as an integer linear program generates parametric, multi-scenario schedule table-driven execution. By inspecting at runtime, our online scheduler...
Due to their natural and inherent way of capturing concurrency, dataflow descriptions stream programs have seen prevalent usage in fields such as video processing, networks scientific computing. They are often deployed on manycore, heterogeneous distributed architectures. Despite robust research the topic, obstacles still exist evaluating performance accurately, especially without a complete implementation down selected platforms. In this work we introduce provide proof concept for an...
Stream processing applications are naturally described as dataflow programs. Dataflow programs modelled actor networks well suited to describe concurrent and computationally intensive problems. Realistic typically characterized by highly dynamic behaviour, limiting the applicability of static analysis techniques. In this work we explore using analyses making use causation traces; graphs which capture instances program's execution. We outline how they can be used inform pipelining...