GPUdmm: A high-performance and memory-oblivious GPU architecture using dynamic memory management

Youngsok Kim, Jaewon Lee, Jae Eon Jo, Jangwoo Kim

Research output: Chapter in Book/Report/Conference proceedingConference contribution

20 Citations (Scopus)

Abstract

GPU programmers suffer from programmer-managed GPU memory because both performance and programmability heavily depend on GPU memory allocation and CPU-GPU data transfer mechanisms. To improve performance and programmability, programmers should be able to place only the data frequently accessed by GPU on GPU memory while overlapping CPU-GPU data transfers and GPU executions as much as possible. However, current GPU architectures and programming models blindly place entire data on GPU memory, requiring a significantly large GPU memory size. Otherwise, they must trigger unnecessary CPU-GPU data transfers due to an insufficient GPU memory size. In this paper, we propose GPUdmm, a novel GPU architecture to enable high-performance and memory-oblivious GPU programming. First, GPUdmm uses GPU memory as a cache of CPU memory to provide programmers a view of the CPU memory-sized programming space. Second, GPUdmm achieves high performance by exploiting data locality and dynamically transferring data between CPU and GPU memories while effectively overlapping CPU-GPU data transfers and GPU executions. Third, GPUdmm can further reduce unnecessary CPU-GPU data transfers by exploiting simple programmer hints. Our carefully designed and validated experiments (e.g., PCIe/DMA timing) against representative benchmarks show that GPUdmm can achieve up to five times higher performance for the same GPU memory size, or reduce the GPU memory size requirement by up to 75% while maintaining the same performance.

Original languageEnglish
Title of host publication20th IEEE International Symposium on High Performance Computer Architecture, HPCA 2014
PublisherIEEE Computer Society
Pages546-557
Number of pages12
ISBN (Print)9781479930975
DOIs
Publication statusPublished - 2014
Event20th IEEE International Symposium on High Performance Computer Architecture, HPCA 2014 - Orlando, FL, United States
Duration: 2014 Feb 152014 Feb 19

Publication series

NameProceedings - International Symposium on High-Performance Computer Architecture
ISSN (Print)1530-0897

Other

Other20th IEEE International Symposium on High Performance Computer Architecture, HPCA 2014
CountryUnited States
CityOrlando, FL
Period14/2/1514/2/19

All Science Journal Classification (ASJC) codes

  • Hardware and Architecture

Fingerprint Dive into the research topics of 'GPUdmm: A high-performance and memory-oblivious GPU architecture using dynamic memory management'. Together they form a unique fingerprint.

Cite this