Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


Kian_Nejadfard_thesis_formatfix.pdf (265.67 KB)

ETD Abstract Container

Abstract Header

Context-aware automated refactoring for unified memory allocation in NVIDIA CUDA programs

Nejadfard, Kian
http://rave.ohiolink.edu/etdc/view?acc_num=csu1624622944458295

Abstract Details

2021, Master of Computer and Information Science, Cleveland State University, Washkewicz College of Engineering.
Refactoring tools, whether automatic or semi-automatic, are an integral part of the software development life cycle. Software libraries and frameworks evolve over time. Programs that use them should also evolve over time, or risk falling behind on compatibility with modern advancements. In the case of NVIDIA CUDA's platform for general-purpose GPU programming, using the more modern unified memory architecture allows for simplifying program source code, reducing bugs associated with manual memory management between host and device memory, and benefiting from memory transfer optimization via automatic handling of memory. This thesis proposes the development of a refactoring tool to handle this transition automatically, therefore freeing developers from the burden and risks involved in manually refactoring large code bases.
Janche Sang (Advisor)
Janche Sang (Committee Chair)
Haodong Wang (Committee Member)
Jingru Zhang (Committee Member)
Computer Engineering; Computer Science
NVIDIA CUDA; NVIDIA; CUDA; refactoring; automated refactoring; context-aware; unified memory; managed memory; source-to-source translation; LLVM; Clang

Recommended Citations

Citations

  • Nejadfard, K. (2021). Context-aware automated refactoring for unified memory allocation in NVIDIA CUDA programs [Master's thesis, Cleveland State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=csu1624622944458295

    APA Style (7th edition)

  • Nejadfard, Kian. Context-aware automated refactoring for unified memory allocation in NVIDIA CUDA programs. 2021. Cleveland State University, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=csu1624622944458295.

    MLA Style (8th edition)

  • Nejadfard, Kian. "Context-aware automated refactoring for unified memory allocation in NVIDIA CUDA programs." Master's thesis, Cleveland State University, 2021. http://rave.ohiolink.edu/etdc/view?acc_num=csu1624622944458295

    Chicago Manual of Style (17th edition)

csu1624622944458295
639
© 2021, all rights reserved.
This open access ETD is published by Cleveland State University and OhioLINK.