Global ETD Search

Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.

Results

Showing 1 to 20 of 54 for “"work stealing"”.

  1. Optimizing work stealing algorithms with scheduling constraints

    … has gained popularity in conjunction with work-stealing schedulers. Random work-stealing schedulers have been shown to effectively perform dynamic load balancing, yielding provably-efficient schedules and space bounds on shared-memory architectures with uniform memory models. However, the …

    uiuc Repository record for Optimizing work stealing algorithms with scheduling constraints (opens in a new tab)

  2. Compiler support for work-stealing parallel runtime systems

    … and Fortress. Scheduling algorithms based on work-stealing, as embodied in Cilk's implementation of dynamic spawn-sync parallelism, are gaining in popularity but also have inherent limitations. In this thesis, we focus on the compiler support needed to extend work-stealing for dynamic …

    rice Repository record for Compiler support for work-stealing parallel runtime systems (opens in a new tab)

  3. Work-Stealing Scheduler for Parallel Cache-Adaptive Algorithms

    … We then introduce a scheduler that leverages work-stealing strategies while respecting the constraints of cache adaptivity. Finally, we validate our approach by applying this scheduler to specific algorithms, demonstrating its ability to support optimal cache usage and efficient computation in …

    wustl Repository record for Work-Stealing Scheduler for Parallel Cache-Adaptive Algorithms (opens in a new tab)

  4. Embracing Explicit Communication in Work-Stealing Runtime Systems

    … abstraction is that of a task, a piece of work, usually a function or a closure, that is safe to execute in parallel with other tasks. Scheduling decisions, including the mapping of tasks to threads, are made by the runtime system and are not imposed on the programmer. Tasks are well-suited …

    bayreuth Repository record for Embracing Explicit Communication in Work-Stealing Runtime Systems (opens in a new tab)

  5. Escalonamento Work-Stealing de programas Divisão-e-Conquista com MPI-2

    … programação Divisão-e-Conquista e uma estratégia Work-Stealing para balancear os processos MPI, em ambientes heterogêneos e/ou dinâmicos, em tempo de execução. Este trabalho explica como implementar uma aplicação segundo o modelo de Divisão-e-Conquista com MPI, bem como a implementação de uma …

    brazil-ufrgs Repository record for Escalonamento Work-Stealing de programas Divisão-e-Conquista com MPI-2 (opens in a new tab)

  6. A Scalable Locality-aware Adaptive Work-stealing Scheduler for Multi-core Task Parallelism

    … to the runtime scheduler. Our thesis is that work-stealing schedulers with adaptive scheduling policies and locality-awareness can provide a scalable and robust runtime foundation for multicore task parallelism. We evaluate our thesis using the new Scalable Locality-aware Adaptive …

    rice Repository record for A Scalable Locality-aware Adaptive Work-stealing Scheduler for Multi-core Task Parallelism (opens in a new tab)

  7. A Work-Stealing For Dynamic Workload Balancing On CPU-GPU Heterogeneous Computing Platforms

    Although many general purpose workloads have been accelerated on graphical processing units (gpus) over the last decade, other applications whose runtime behaviors are dynamic and irregular such as ones based on trees and graphs have suffered from serious workload imbalance problem caused by …

    mississippi Repository record for A Work-Stealing For Dynamic Workload Balancing On CPU-GPU Heterogeneous Computing Platforms (opens in a new tab)

  8. Optimized Event-Driven Runtime Systems for Programmability and Performance

    … machine and sacrifice portability. In our former work, we proposed the Data-Driven Tasks model, which constitutes expressive and portable parallelism by only requiring the application programmer to declare the inherent dependences of the application. In this work, we observe another instantiation …

    rice Repository record for Optimized Event-Driven Runtime Systems for Programmability and Performance (opens in a new tab)

  9. Enabling Moldability in OpenMP

    … threads should run tasks, it often uses a simple work-stealing scheduler as they evenly distribute tasks among cores. This is the method used by LLVMs OpenMP runtime. But today, HPC systems often consist of multiple sockets, each with many cores and non-uniform memory access (NUMA). This creates a …

    chalmers Repository record for Enabling Moldability in OpenMP (opens in a new tab)

  10. Parallelization in Rust with fork-join and friends: Creating the fork-join framework

    … the design, implementation and benchmarking of a work stealing fork-join library, called ForkJoin, for the new language Rust. Rust is a programming language with a novel approach to memory safety and concurrency, and guarantees memory safety through zero-cost abstractions and thorough checks at …

    chalmers Repository record for Parallelization in Rust with fork-join and friends: Creating the fork-join framework (opens in a new tab)

  11. Parallel, lock-free framework for decision diagram (DD) based search

    … programming and introduces a new lock-free work-stealing algorithm designed to overcome its scheduling bottlenecks. Existing work-stealing queues performed inefficiently for solver workloads involving bulk node generation and single-stealer concurrency. To address this, an unbounded queue …

    iastate Repository record for Parallel, lock-free framework for decision diagram (DD) based search (opens in a new tab)

  12. A runtime system for data-flow task programming on multicore architectures with accelerators

    … for asynchronous execution and scheduling. Work stealing with heuristics showed significant performance results, but did not consider the computing power of different resources. Next, we designed a scheduling framework and a performance model to support scheduling strategies over XKaapi …

    brazil-ufrgs Repository record for A runtime system for data-flow task programming on multicore architectures with accelerators (opens in a new tab)

  13. Dynamic Load Balancing and Elasticity in a Distributed Heterogeneous Task-based Dataflow Runtime

    … this paradigm, individual processes divide the workload between nodes, and threads further divide the tasks within each node. This model proved efficient when hardware was homogeneous, and applications were regular and balanced. However, with the advent of hardware heterogeneity and increased …

    anu Repository record for Dynamic Load Balancing and Elasticity in a Distributed Heterogeneous Task-based Dataflow Runtime (opens in a new tab)

  14. ILAN: The Interference- and Locality-Aware NUMA Scheduler

    … to the default OpenMP tasking scheduler, the work stealing scheduler, yielding improved data locality. Furthermore, the ILAN scheduler utilizes moldability to incorporate interference awareness, dynamically reducing the number of OpenMP threads to mitigate the effects of interference while …

    chalmers Repository record for ILAN: The Interference- and Locality-Aware NUMA Scheduler (opens in a new tab)

  15. High-level execution models for multicore architectures

    … programmability. Second, we propose the use of work-stealing runtimes for coarse-grained dataflow parallelism and doacross runtimes for fine-grained dataflow parallelism, both of which can be expressed uniformly in DFGR. We justify this approach by demonstrating the performance of DFGR on two …

    rice Repository record for High-level execution models for multicore architectures (opens in a new tab)

  16. Workload scheduling analysis in geophysical numerical methods

    … an application non-scalable. Therefore this work investigates the workload scheduling in geophysical methods. Three types of workload scheduling will be compared: centralized dynamic (CD), decentralized static (DS), and decentralized dynamic (DD). The CD was implemented with the technique …

    brazil-ufrn Repository record for Workload scheduling analysis in geophysical numerical methods (opens in a new tab)

  17. Visualizing algorithmic progression in task-based parallelism : a visual analytics methodology for evaluating dynamic schedulers

    … study contrasts two dynamic scheduling policies: Work Stealing (WS) and Locality Work Stealing (LWS). While macroscopic metrics indicate a significant performance degradation for LWS, they cannot explain its root cause. The k-hills visualization successfully decodes this anomaly, visually proving …

    brazil-ufrgs Repository record for Visualizing algorithmic progression in task-based parallelism : a visual analytics methodology for evaluating dynamic schedulers (opens in a new tab)

  18. EFFICIENT FINE-GRAIN COOPERATIVE EXECUTION OF DYNAMIC TASK PARALLELISM ON HETEROGENEOUS MULTI/MANYCORE SYSTEMS

    … when abundant explicit parallelism exists in the workloads. However, large-scale manycore processors often lack hardware-based cache co- herence. There is a growing trend towards a tighter integration between multicore and manycore processors, forming heterogeneous multi/manycore systems. These …

    cornell Repository record for EFFICIENT FINE-GRAIN COOPERATIVE EXECUTION OF DYNAMIC TASK PARALLELISM ON HETEROGENEOUS MULTI/MANYCORE SYSTEMS (opens in a new tab)

  19. Runtime Approaches to Improve the Efficiency of Hybrid and Irregular Applications

    … operations. We combine gang-scheduling with work-stealing for data- parallel tasks with frequent inter/intra-node communication in the task graphs so as to reduce interference and expensive context switching operations. We also propose improved victim selection policies for work-stealing to …

    gatech Repository record for Runtime Approaches to Improve the Efficiency of Hybrid and Irregular Applications (opens in a new tab)

  20. High Performance and Dependable Asynchronous Communication on Multi-Core Systems

    … such as operating systems, databases, networks, and language runtimes, enabling data transfer, task distribution, and component decoupling. However, with the waning of Moore's Law and the rise of heterogeneous multi-core architectures, existing methods face challenges, including …

    qucosa-diss

Page 1 of 3