document

Programming tensor cores from an image processing DSL

Sioutas, S. (author), Stuijk, S. (author), Basten, T. (author), Somers, L. (author), Corporaal, H. (author)

Tensor Cores (TCUs) are specialized units first introduced by NVIDIA in the Volta microarchitecture in order to accelerate matrix multiplications for deep learning and linear algebra workloads. While these units have proved to be capable of providing significant speedups for specific applications, their programmability remains difficult for the...

conference paper 2020

document

Scenarios in the design of flexible manufacturing systems

Basten, T. (author), Bastos, J. (author), Medina, R. (author), van der Sanden, B. (author), Geilen, M.C.W. (author), Goswami, D. (author), Reniers, M.A. (author), Stuijk, S. (author), Voeten, J.P.M. (author)

bookPart 2020

document

Scenarios in dataflow modeling and analysis

Geilen, M.C.W. (author), Skelin, M. (author), van Kampenhout, J.R. (author), Ara, H.A. (author), Basten, T. (author), Stuijk, S. (author), Goossens, K.G.W. (author)

bookPart 2020

document

Reconfigurable pipelined control systems

Sanchez, R.M. (author), Nikkhah, S.T. (author), Goswami, D. (author), Heemels, W. (author), Stuijk, S. (author), Basten, T. (author)

Pipelining is a promising technique to deal with long delays in control loops, for instance due to intensive data processing. Reconfigurable pipelined control (RPC) is of particular interest, as it has an improved performance compared to a static configuration. We present design and implementation techniques for RPC systems.

article 2020

document

Schedule Synthesis for Halide Pipelines on GPUs

Sioutas, S. (author), Stuijk, S. (author), Basten, T. (author), Corporaal, H. (author), Somers, L. (author)

The Halide DSL and compiler have enabled high-performance code generation for image processing pipelines targeting heterogeneous architectures through the separation of algorithmic description and optimization schedule. However, automatic schedule generation is currently only possible for multi-core CPU architectures. As a result, expert...

conference paper 2020

document

Designing a controller with image-based pipelined sensing and additive uncertainties

Medina, R. (author), Valencia, J. (author), Stuijk, S. (author), Goswami, D. (author), Basten, T. (author)

Pipelined image-based control uses parallel instances of its image-processing algorithm in a pipelined fashion to improve the quality of control. A performance-oriented control design improves the controller settling time with each additional processing resource, which creates a resources-performance trade-off. In real-life applications, it is...

article 2019

document

Implementation-aware design of image-based control with on-line measurable variable-delay

Medina Sánchez, R.A. (author), Stuijk, S. (author), Goswami, D. (author), Basten, A.A. (author)

Image-based control uses image-processing algorithms to acquire sensing information. The sensing delay associated with the image-processing algorithm is typically platform-dependent and time-varying. Modern embedded platforms allow to characterize the sensing delay at design-time obtaining a delay histogram, and at run-time measuring its precise...

conference paper 2019

document

Schedule synthesis for halide pipelines through reuse analysis

Sioutas, S. (author), Stuijk, S. (author), Waeijen, L. (author), Basten, T. (author), Corporaal, H. (author), Somers, L. (author)

Efficient code generation for image processing applications continues to pose a challenge in a domain where high performance is often necessary to meet real-time constraints. The inherently complex structure found in most image-processing pipelines, the plethora of transformations that can be applied to optimize the performance of an...

article 2019

document

Loop transformations leveraging hardware prefetching

Sioutas, S. (author), Stuijk, S. (author), Corporaal, H. (author), Basten, A.A. (author), Somers, L. (author)

Memory-bound applications heavily depend on the bandwidth of the system in order to achieve high performance. Improving temporal and/or spatial locality through loop transformations is a common way of mitigating this dependency. However, choosing the right combination of optimizations is not a trivial task, due to the fact that most of them...

conference paper 2018

document

Identifying bottlenecks in manufacturing systems using stochastic criticality analysis

Bastos, J. (author), van der Sanden, B. (author), Donk, O. (author), Voeten, J. (author), Stuijk, S. (author), Schiffelers, R. (author), Corporaal, H. (author)

System design is a difficult process with many design-choices for which the impact may be difficult to foresee. Manufacturing system design is no exception to this. Increased use of flexible manufacturing systems which are able to perform different operations/use-cases further raises the design complexity. One important criterion to consider is...

conference paper 2018

document

Effective link quality estimation as a means to improved end-to-end packet delivery in high traffic mobile ad hoc networks

Afzal, S.R. (author), Stuijk, S. (author), Nabi, M. (author), Basten, T. (author)

Accurate link quality estimation is a fundamental building block in quality aware multi hop routing. In an inherently lossy, unreliable and dynamic medium such as wireless, the task of accurate estimation becomes very challenging. Over the years ETX has been widely used as a reliable link quality estimation metric. However, more recently it has...

article 2017

document

Mapping of synchronous dataflow graphs on MPSoCs based on parallelism enhancement

Tang, Q. (author), Basten, T. (author), Geilen, M. (author), Stuijk, S. (author), Wei, J.B. (author)

Multi-processor systems-on-chips are widely adopted in implementing modern streaming applications to satisfy the ever increasing computation requirements. To take advantage of this kind of platform, it is necessary to map tasks of the application properly to different processors, so as to fully exploit the inherent task-level parallelism and...

article 2017

document

Exploring the trade-off between processing resources and settling time in image-based control through LQR tuning

Medina, R. (author), Stuijk, S. (author), Goswami, D. (author), Basten, T. (author)

Image-Based control systems extract information by a camera and an image processing algorithm. The challenge of such controllers is that the sensing latency deteriorates the control performance. Multi-core technology can be used to implement the sensing algorithm in a pipelined fashion. More processing resources potentially lead to better...

conference paper 2017

document

Task-FIFO co-scheduling of streaming applications on MPSoCs with predictable memory hierarchy

Tang, Q. (author), Basten, T. (author), Geilen, M. (author), Stuijk, S. (author), Wei, J.B. (author)

This article studies the scheduling of real-time streaming applications on multiprocessor systems-on-chips with predictable memory hierarchy. An iteration-based task-FIFO co-scheduling framework is proposed for this problem. We obtain FIFO size distributions using Pareto space searching, based on which the task-toprocessor mapping is obtained...

article 2017

document

Performance analysis of weakly-consistent scenario-aware dataflow graphs

Geilen, M. (author), Falk, J. (author), Haubelt, C. (author), Basten, T. (author), Theelen, B. (author), Stuijk, S. (author)

The timed dataflow model of computation is a useful performance analysis tool for electronic system level design automation and embedded software synthesis. Its determinism gives it strong analyzability properties. Its monotonic temporal behavior provides hard real-time guarantees on throughput and latency. It is expressive enough to cover a...

article 2017

document

Reconfigurable pipelined sensing for image-based control

Medina, R. (author), Stuijk, S. (author), Goswami, D. (author), Basten, T. (author)

Image-based control systems are becoming common in domains such as robotics, healthcare and industrial automation. Coping with a long sample period because of the latency of the image processing algorithm is an open challenge. Modern multi-core platforms allow to address this challenge by pipelining the sensing algorithm. Often, such systems...

conference paper 2016

document

A Fast Estimator of Performance with Respect to the Design Parameters of Self Re-Entrant Flowshops

Waqas, U. (author), Geilen, M. (author), Stuijk, S. (author), Pinxten, J.V. (author), Basten, T. (author), Somers, L. (author), Corporaal, H. (author)

Self re-entrant flowshops consist of machines which process jobs several times. They are found in applications like TFT-LCD assembly, LED manufacturing and industrial printing. The structure of a self re-entrant flowshop influences its performance. To get better performance while reducing costs a fast performance estimation method can be used to...

conference paper 2016

document

INLyD: Inter-network-layer delay as a low-cost quality metric for multi-hop routing in wireless mobile networks

Afzal, S.R. (author), Nabi, M. (author), Stuijk, S. (author), Basten, T. (author)

The need for authentic and effective portrayal of the spatiotemporally changing quality of wireless links has gained wide attention especially over the last decade. Software-based link quality estimators (LQE) classify links with help of packet reception ratio (PRR), required number of packet transmissions (RNP) and scoring/grading schemes that...

conference paper 2016

document

Multiconstraint Static Scheduling of Synchronous Dataflow Graphs Via Retiming and Unfolding

Zhu, X.Y. (author), Geilen, M. (author), Basten, T. (author), Stuijk, S. (author)

Synchronous dataflow graphs (SDFGs) are widely used to represent digital signal processing algorithms and streaming media applications. This paper presents several methods for binding and scheduling SDFGs on a multiprocessor platform. Exploring the state space generated by a self-timed execution (STE) of an SDFG, we present an exact method for...

article 2016

document

Robust online face tracking-by-detection

Comaschi, F. (author), Stuijk, S. (author), Basten, T. (author), Corporaal, H. (author)

The problem of online face tracking from unconstrained videos is still unresolved. Challenges range from coping with severe online appearance variations to coping with occlusion. We propose RFTD (Robust Face Tracking-by-Detection), a system which combines tracking and detection into a single framework to robustly track a face from unconstrained...

conference paper 2016