Matching Items (39)
Description
Stream computing has emerged as an importantmodel of computation for embedded system applications particularly in the multimedia and network processing domains. In recent past several programming languages and embedded multi-core processors have been proposed for streaming applications. This thesis examines the execution and dynamic scheduling of stream programs on embedded multi-core processors. The thesis addresses the problem in the context of a multi-tasking environment with a time varying allocation of processing elements for a particular streaming application. As a solution the thesis proposes a two step approach where the stream program is compiled to gather key application information, and to generate re-targetable code. A light weight dynamic scheduler incorporates the second stage of the approach. The dynamic scheduler utilizes the static information and available resources to assign or partition the application across the multi-core architecture. The objective of the dynamic scheduler is to maximize the throughput of the application, and it is sensitive to the resource (processing elements, scratch-pad memory, DMA bandwidth) constraints imposed by the target architecture. We evaluate the proposed approach by compiling and scheduling benchmark stream programs on a representative embedded multi-core processor. We present experimental results that evaluate the quality of the solutions generated by the proposed approach by comparisons with existing techniques.
ContributorsLee, Haeseung (Author) / Chatha, Karamvir (Thesis advisor) / Vrudhula, Sarma (Committee member) / Chakrabarti, Chaitali (Committee member) / Wu, Carole-Jean (Committee member) / Arizona State University (Publisher)
Created2013
Description
Rapid technology scaling, the main driver of the power and performance improvements of computing solutions, has also rendered our computing systems extremely susceptible to transient errors called soft errors. Among the arsenal of techniques to protect computation from soft errors, Control Flow Checking (CFC) based techniques have gained a reputation of effective, yet low-cost protection mechanism. The basic idea is that, there is a high probability that a soft-fault in program execution will eventually alter the control flow of the program. Therefore just by making sure that the control flow of the program is correct, significant protection can be achieved. More than a dozen techniques for CFC have been developed over the last several decades, ranging from hardware techniques, software techniques, and hardware-software hybrid techniques as well. Our analysis shows that existing CFC techniques are not only ineffective in protecting from soft errors, but cause additional power and performance overheads. For this analysis, we develop and validate a simulation based experimental setup to accurately and quantitatively estimate the architectural vulnerability of a program execution on a processor micro-architecture. We model the protection achieved by various state-of-the-art CFC techniques in this quantitative vulnerability estimation setup, and find out that software only CFC protection schemes (CFCSS, CFCSS+NA, CEDA) increase system vulnerability by 18% to 21% with 17% to 38% performance overhead. Hybrid CFC protection (CFEDC) increases vulnerability by 5%, while the vulnerability remains almost the same for hardware only CFC protection (CFCET); notwithstanding the hardware overheads of design cost, area, and power incurred in the hardware modifications required for their implementations.
ContributorsRhisheekesan, Abhishek (Author) / Shrivastava, Aviral (Thesis advisor) / Colbourn, Charles Joseph (Committee member) / Wu, Carole-Jean (Committee member) / Arizona State University (Publisher)
Created2013
Description
With increasing transistor volume and reducing feature size, it has become a major design constraint to reduce power consumption also. This has given rise to aggressive architectural changes for on-chip power management and rapid development to energy efficient hardware accelerators. Accordingly, the objective of this research work is to facilitate software developers to leverage these hardware techniques and improve energy efficiency of the system. To achieve this, I propose two solutions for Linux kernel: Optimal use of these architectural enhancements to achieve greater energy efficiency requires accurate modeling of processor power consumption. Though there are many models available in literature to model processor power consumption, there is a lack of such models to capture power consumption at the task-level. Task-level energy models are a requirement for an operating system (OS) to perform real-time power management as OS time multiplexes tasks to enable sharing of hardware resources. I propose a detailed design methodology for constructing an architecture agnostic task-level power model and incorporating it into a modern operating system to build an online task-level power profiler. The profiler is implemented inside the latest Linux kernel and validated for Intel Sandy Bridge processor. It has a negligible overhead of less than 1\% hardware resource consumption. The profiler power prediction was demonstrated for various application benchmarks from SPEC to PARSEC with less than 4\% error. I also demonstrate the importance of the proposed profiler for emerging architectural techniques through use case scenarios, which include heterogeneous computing and fine grained per-core DVFS. Along with architectural enhancement in general purpose processors to improve energy efficiency, hardware accelerators like Coarse Grain reconfigurable architecture (CGRA) are gaining popularity. Unlike vector processors, which rely on data parallelism, CGRA can provide greater flexibility and compiler level control making it more suitable for present SoC environment. To provide streamline development environment for CGRA, I propose a flexible framework in Linux to do design space exploration for CGRA. With accurate and flexible hardware models, fine grained integration with accurate architectural simulator, and Linux memory management and DMA support, a user can carry out limitless experiments on CGRA in full system environment.
ContributorsDesai, Digant Pareshkumar (Author) / Vrudhula, Sarma (Thesis advisor) / Chakrabarti, Chaitali (Committee member) / Wu, Carole-Jean (Committee member) / Arizona State University (Publisher)
Created2013
Description
The Mobile Waterway Monitor seeks to monitor water in an unexplored way. The module is buoyant and will float with the current as well as harvests solar energy. In short, the Mobile Waterway Monitor excels in size constraints, flexibility, extensibility, and capability. This current following monitor can show both measured trends like pH and interpolated trends like water speed, river contours, and elevation drop. The MWM strikes a balance between accuracy, portability, and being multi-purpose.
ContributorsStribrny, Kody John (Author) / Vrudhula, Sarma (Thesis director) / Wu, Carole-Jean (Committee member) / Computer Science and Engineering Program (Contributor) / Barrett, The Honors College (Contributor)
Created2017-05
Description
Arc Routing Problems (ARPs) are a type of routing problem that finds routes of minimum total cost covering the edges or arcs in a graph representing street or road networks. They find application in many essential services such as residential waste collection, winter gritting, and others. Being NP-hard, solutions are usually found using heuristic methods. This dissertation contributes to heuristics for ARP, with a focus on the Capacitated Arc Routing Problem (CARP) with additional constraints. In operations such as residential waste collection, vehicle breakdown disruptions occur frequently. A new variant Capacitated Arc Re-routing Problem for Vehicle Break-down (CARP-VB) is introduced to address the need to re-route using only remaining vehicles to avoid missing services. A new heuristic Probe is developed to solve CARP-VB. Experiments on benchmark instances show that Probe is better in reducing the makespan and hence effective in reducing delays and avoiding missing services.
In addition to total cost, operators are also interested in solutions that are attractive, that is, routes that are contiguous, compact, and non-overlapping to manage the work. Operators may not adopt a solution that is not attractive even if it is optimum. They are also interested in solutions that are balanced in workload to meet equity requirements. A new multi-objective memetic algorithm, MA-ABC is developed, that optimizes three objectives: Attractiveness, makespan, and total cost. On testing with benchmark instances, MA-ABC was found to be effective in providing attractive and balanced route solutions without affecting the total cost.
Changes in the problem specification such as demand and topology occurs frequently in business operations. Machine learning be applied to learn the distribution behind these changes and generate solutions quickly at time of inference. Splice is a machine learning framework for CARP that generates closer to optimum solutions quickly using a graph neural network and deep Q-learning. Splice can solve several variants of node and arc routing problems using the same architecture without any modification. Splice was trained and tested using randomly generated instances. Splice generated solutions faster that are also better in comparison to popular metaheuristics.
ContributorsRamamoorthy, Muhilan (Author) / Syrotiuk, Violet R. (Thesis advisor) / Forrest, Stephanie (Committee member) / Mirchandani, Pitu (Committee member) / Sen, Arunabha (Committee member) / Arizona State University (Publisher)
Created2022
Description
Coffee leaf rust (CLR) is an aggressive disease that has caused devastating production losses around the world. While coffee productivity under shade conditions has been simulated in sufficient detail, explicit modeling of the interactions between shading levels, microclimate, coffee plant physiology, and CLR progression remains largely unexplored, despite the recognized influence of shade on CLR epidemics. This dissertation introduces a new model, SpatialRust, as an initial approximation to an integrative simulation framework where farm design and management strategies can be linked to coffee production and CLR epidemic outcomes. SpatialRust considers stylized processes describing the dynamics of shade trees, coffee plants, and CLR and their interactions within a spatially explicit context. The dissertation then presents three experiments conducted using SpatialRust simulations. The first experiment investigates the role of shading as a mitigation tool for CLR outbreaks. It demonstrates that shade can effectively mitigate CLR epidemics when the conditions are otherwise conducive to a major CLR epidemic. Additionally, the experiment reveals complex effects of different shade management approaches, underscoring the potential value of future empirical studies that consider temporal and spatial shade dynamics in relation to CLR outcomes. The second experiment focuses on the financial balance of farms and examines how farmer preferences and needs influence farm management strategies. The findings indicate that incorporating CLR mitigation as part of the strategy's goals leads to positive long-term farm performance, even when planning for the short term. In the third experiment, the scope of the simulations is expanded to include neighboring coffee farms. The results demonstrate that the strategies adopted by immediate neighbors can affect the performance of individual farms, emphasizing the importance of considering the broader coffee-growing landscape context. This work shows that the integration of farm management practices and the resulting shading effects into a spatially explicit framework can provide valuable insights into the dynamics of CLR epidemics and how they respond to farmers' decisions.
ContributorsVanegas Ferro, Manuela (Author) / Janssen, Marcus A. (Thesis advisor) / Forrest, Stephanie (Committee member) / Restrepo Restrepo, Silvia (Committee member) / Arizona State University (Publisher)
Created2023
Description
Graphic Processing Units (GPUs) have become a key enabler of the big-data revolution, functioning as defacto co-processors to accelerate large-scale computation. As the GPU programming stack and tool support have matured, the technology has alsobecome accessible to programmers. However, optimizing programs to run efficiently on GPUs requires developers to have both detailed understanding of the application logic and significant knowledge of parallel programming and GPU architectures.
This dissertation proposes GEVO, a tool for automatically tuning the performance of GPU kernels in the LLVM representation to meet desired criteria. GEVO uses population-based search to find edits to programs compiled to LLVM-IR which improves performance on desired criteria and retains required functionality. The evaluation of GEVO on the Rodinia benchmark suite demonstrates many runtime optimization techniques. One of the key insights is that semantic relaxation enables GEVO to
discover these optimizations that are usually prohibited by the compiler. GEVO also explores many other optimizations, including architecture- and application-specific. A follow-up evaluation of three bioinformatics applications at their different stages of development suggests that GEVO can optimize programs as well as human experts, sometimes even into a code base that is beyond a programmer’s reach. Furthermore, to unshackle the constraint of GEVO in optimizing neural network (NN) models,
GEVO-ML is proposed by extending the representation capability of GEVO, where NN models and the training/prediction process are uniformly represented in a single intermediate language. An evaluation of GEVO-ML shows that GEVO-ML can
optimize network models similar to how human developers improve model design, for example, by changing learning rates or pruning non-essential parameters. These results showcase the potential of automated program optimization tools to both reduce
the optimization burden for researchers and provide new insights for GPU experts.
ContributorsLiou, Jhe-Yu (Author) / Forrest, Stephanie (Thesis advisor) / Wu, Carole-Jean (Thesis advisor) / Lee, Yann-Hang (Committee member) / Weimer, Westley (Committee member) / Arizona State University (Publisher)
Created2023
Modeling Spatial Competition and Adaptive Therapy Protocols in Three-Dimensional Vascularized Tumors
Description
In contrast to traditional chemotherapy for cancer which fails to address tumor heterogeneity, raises patients’ levels of toxicity, and selects for drug-resistant cells, adaptive therapy applies ideas from cancer ecology in employing low-dose drugs to encourage competition between cancerous cells, reducing toxicity and potentially prolonging disease progression. Despite promising results in some clinical trials, optimizing adaptive therapy routines involves navigating a vast space of combina- torial possibilities, including the number of drugs, drug holiday duration, and drug dosages. Computational models can serve as precursors to efficiently explore this space, narrowing the scope of possibilities for in-vivo and in-vitro experiments which are time-consuming, expensive, and specific to tumor types. Among the existing modeling techniques, agent-based models are particularly suited for studying the spatial inter- actions critical to successful adaptive therapy. In this thesis, I introduce CancerSim, a three-dimensional agent-based model fully implemented in C++ that is designed to simulate tumorigenesis, angiogenesis, drug resistance, and resource competition within a tissue. Additionally, the model is equipped to assess the effectiveness of various adaptive therapy regimens. The thesis provides detailed insights into the biological motivation and calibration of different model parameters. Lastly, I propose a series of research questions and experiments for adaptive therapy that CancerSim can address in the pursuit of advancing cancer treatment strategies.
ContributorsShah, Sanjana Saurin (Author) / Daymude, Joshua J (Thesis advisor) / Forrest, Stephanie (Committee member) / Maley, Carlo C (Committee member) / Arizona State University (Publisher)
Created2023
Description
Resistance to existing anti-cancer drugs poses a key challenge in the field of medical oncology, in that it results in the tumor not responding to treatment using the same medications to which it responded previously, leading to treatment failure. Adaptive therapy utilizes evolutionary principles of competitive suppression, leveraging competition between drug resistant and drug sensitive cells, to keep the population of drug resistant cells under control, thereby extending time to progression (TTP), relative to standard treatment using maximum tolerated dose (MTD). Development of adaptive therapy protocols is challenging, as it involves many parameters, and the number of parameters increase exponentially for each additional drug. Furthermore, the drugs could have a cytotoxic (killing cells directly), or a cytostatic (inhibiting cell division) mechanism of action, which could affect treatment outcome in important ways. I have implemented hybrid agent-based computational models to investigate adaptive therapy, using either a single drug (cytotoxic or cytostatic), or two drugs (cytotoxic or cytostatic), simulating three different adaptive therapy protocols for treatment using a single drug (dose modulation, intermittent, dose-skipping), and seven different treatment protocols for treatment using two drugs: three dose modulation (DM) protocols (DM Cocktail Tandem, DM Ping-Pong Alternate Every Cycle, DM Ping-Pong on Progression), and four fixed-dose (FD) protocols (FD Cocktail Intermittent, FD Ping-Pong Intermittent, FD Cocktail Dose-Skipping, FD Ping-Pong Dose-Skipping). The results indicate a Goldilocks level of drug exposure to be optimum, with both too little and too much drug having adverse effects. Adaptive therapy works best under conditions of strong cellular competition, such as high fitness costs, high replacement rates, or high turnover. Clonal competition is an important determinant of treatment outcome, and as such treatment using two drugs leads to more favorable outcome than treatment using a single drug. Switching drugs every treatment cycle (ping-pong) protocols work particularly well, as well as cocktail dose modulation, particularly when it is feasible to have a highly sensitive measurement of tumor burden. In general, overtreating seems to have adverse survival outcome, and triggering a treatment vacation, or stopping treatment sooner when the tumor is shrinking seems to work well.
ContributorsSaha, Kaushik (Author) / Maley, Carlo C (Thesis advisor) / Forrest, Stephanie (Committee member) / Anderson, Karen S (Committee member) / Cisneros, Luis H (Committee member) / Arizona State University (Publisher)
Created2023
Description
Nucleic acid nanotechnology is a field of nanoscale engineering where the sequences of deoxyribonucleicacid (DNA) and ribonucleic acid (RNA) molecules are carefully designed to create self–assembled nanostructures with higher spatial resolution than is available to top–down fabrication methods. In the 40 year history
of the field, the structures created have scaled from small tile–like structures constructed from a few hundred
individual nucleotides to micron–scale structures assembled from millions of nucleotides using the technique
of “DNA origami”. One of the key drivers of advancement in any modern engineering field is the parallel
development of software which facilitates the design of components and performs in silico simulation of the
target structure to determine its structural properties, dynamic behavior, and identify defects. For nucleic acid
nanotechnology, the design software CaDNAno and simulation software oxDNA are the most popular choices
for design and simulation, respectively. In this dissertation I will present my work on the oxDNA software
ecosystem, including an analysis toolkit, a web–based graphical interface, and a new molecular visualization
tool which doubles as a free–form design editor that covers some of the weaknesses of CaDNAno’s lattice–based design paradigm. Finally, as a demonstration of the utility of these new tools I show oxDNA simulation
and subsequent analysis of a nanoscale leaf–spring engine capable of converting chemical energy into dynamic motion. OxDNA simulations were used to investigate the effects of design choices on the behavior of
the system and rationalize experimental results.
ContributorsPoppleton, Erik (Author) / Sulc, Petr (Thesis advisor) / Yan, Hao (Committee member) / Forrest, Stephanie (Committee member) / Stephanopoulos, Nicholas (Committee member) / Arizona State University (Publisher)
Created2022