Georgia Institute of Technology
Integrating Distributed and Task-Based Shared Memory Parallelism
Abstract
dc:description.abstractThis work presents 3 approaches to representing high-performance, highly-parallel applications at the library/runtime level with a focus on programmability and performance. First, we demonstrate an approach to combining task-based/ULT parallelism with MPI using a previously published continuations-based MPI extension. We then outline a conversion of HClib Selectors—a previously published actor-based PGAS runtime—to Rust to increase its programmability. Finally, we present some general observations about the actor model and an attempt at combining message-passing and threading in a new actor runtime called JacTors.
Degree
thesis:*- Level thesis:degree_level
- Masters
- Department dc:contributor.department
- Computer Science
- Grantor dc:publisher
- Georgia Institute of Technology
- Year dc:date.issued
- 2023
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Parrish, John Thomas Hunter
- Advisor dc:contributor.advisor
-
- Sarkar, Vivek
- Committee members dc:contributor.committeemember
-
- Conte, Tom M.
- Ramachandran, Umakishore
Subjects
dc:subject × 6Rights
- Language dc:language.iso
- en_US
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/1853/72735