Explore chapters and articles related to this topic
Tools and Methodologies for System-Level Design
Published in Louis Scheffer, Luciano Lavagno, Grant Martin, EDA for IC System Design, Verification, and Testing, 2018
Shuvra Bhattacharyya, Wayne Wolf
System-level design is less amenable to synthesis than are logic or physical design. As a result, system-level tools concentrate on modeling, simulation, design space exploration, and design verification. The goal of modeling is to correctly capture the system’s operational semantics, which helps with both implementation and verification. The study of models of computation provides a framework for the description of digital systems. Not only do we need to understand a particular style of computation such as dataflow, but we also need to understand how different models of communication can reliably communicate with each other. Design space exploration tools, such as hardware/software codesign, develop candidate designs to understand trade-offs. Simulation can be used not only to verify functional correctness but also to supply performance and power/energy information for design analysis.
Tools and Methodologies for System-Level Design
Published in Luciano Lavagno, Igor L. Markov, Grant Martin, Louis K. Scheffer, Electronic Design Automation for IC System Design, Verification, and Testing, 2017
Shuvra Bhattacharyya, Marilyn Wolf
System-level design is less amenable to synthesis than are logic or physical design. As a result, system-level tools concentrate on modeling, simulation, design space exploration, and design verification. The goal of modeling is to correctly capture the system’s operational semantics, which helps with both implementation and verification. The study of models of computation provides a framework for the description of digital systems. Not only do we need to understand a particular style of computation, such as dataflow, but we also need to understand how different models of communication can reliably communicate with each other. Design space exploration tools, such as hardware/software codesign, develop candidate designs to understand trade-offs. Simulation can be used not only to verify functional correctness but also to supply performance and power/energy information for design analysis.
NoC and System-Level Design
Published in Hoi-Jun Yoo, Kangmin Lee, Jun Kyoung Kim, Low-Power NoC for High-Performance SoC Design, 2018
Hoi-Jun Yoo, Kangmin Lee, Jun Kyoung Kim
The addition of new behaviors and channels during PE allocation modifies the system model into the architecture model. After PE allocation, the main issue is how to implement the PEs by either software or hardware. Cost functions can be derived from system constraints, and based on the results of the simulation or proper estimation, the PE allocation can be retried. Such simulation or estimation to look for the optimum design is called design space exploration.
Application of machine learning techniques to build digital twins for long train dynamics simulations
Published in Vehicle System Dynamics, 2023
N. Bosso, M. Magelli, R. Trinchero, N. Zampieri
Usually, design space exploration tasks of complex structures are performed by relying on the results of computational expensive simulation experiments via the so-called ‘computational model’. The computational model can be seen as the most accurate synthetic description of the actual behaviour of the system under modelling, able to virtually predict the output quantities of interest for several configurations of the system parameters. Computational models can have several levels of complexity, going from a simple closed-form approximation to the most complicated full-body one. Although full-body simulations are clearly more accurate, they can be extremely expensive in terms of computational costs, thus making design space exploration tasks extremely inefficient when many simulation runs are needed. In the current work, the computational model is represented by the vehicle MB model.
Human-Centered Generative Design Framework: An Early Design Framework to Support Concept Creation and Evaluation
Published in International Journal of Human–Computer Interaction, 2023
H. Onan Demirel, Molly H. Goldstein, Xingang Li, Zhenghui Sha
One should note that we used parametric and non-parametric design approaches when creating cutout geometries for the concept pillars with see-through holes. The concept pillar models included circular, hexagonal, and triangular cutouts. These cutout geometries generated the see-through gaps by removing material from the pillar frame. The concept variants represented under the non-generative category were created using traditional CAD modeling without parametrization. For concept variants created via the GD approach, we parametrized each model, then ran many iterations with varying sizes and distributions using condensers and attractors to mimic a simplified GD modeling approach without the aid of AI. This process created a large pool of concept pillar variants. Then, we selected candidate pillar models that promoted better visibility and carried them into DHM analysis to quantify percent visibility obscuration. The parametric design process differs from the generative design. However, both are subsets of computational design, and the evolution of the theories and applications that fuel the GD development incorporates many elements from the PD approach. The significant difference is that PD does not necessarily involve AI-based algorithms to generate design alternatives automatically. Thus, the design space exploration still depends on human expertise and heuristics. One way to think is that PD is the manual and crude version of GD without AI-based algorithms. In this study, running PD is regarded as the first step towards a more sophisticated GD algorithm and incorporating optimization in the future.
A proximity-based surrogate-assisted method for simulation-based design optimization of a cylinder head water jacket
Published in Engineering Optimization, 2021
Ali Ahrari, Julian Blank, Kalyanmoy Deb, Xianren Li
Various surrogate-assisted algorithms have been proposed in recent years. One of the first multi-objective metamodel-based algorithms is ParEGO (Jones, Schonlau, and Welch 1998), which uses Kriging as the metamodel and decomposes the multi-objective problem into many single-objective problems using the Tchebycheff function (Steuer and Choo 1983). The authors focused on relatively low-dimensional and real-valued functions and assessed the performance after 100 and 250 function evaluations. Wang and Shan (2007) provide an overview of metamodelling metamodelling techniques and their applications to support engineering design optimization. The authors categorize recent development in this research area regarding the needs of design engineers: model approximation, design space exploration and problem formulation.