Performance Evaluation of Advanced Features in CUDA Unified Memory

October 21, 2019 Β· Declared Dead Β· πŸ› Workshop on Memory Centric High Performance Computing

πŸ‘» CAUSE OF DEATH: Ghosted
No code link whatsoever

"No code URL or promise found in abstract"

Evidence collected by the PWNC Scanner

Authors Steven W. D. Chien, Ivy B. Peng, Stefano Markidis arXiv ID 1910.09598 Category cs.DC: Distributed Computing Citations 52 Venue Workshop on Memory Centric High Performance Computing Last Checked 5 months ago
Abstract
CUDA Unified Memory improves the GPU programmability and also enables GPU memory oversubscription. Recently, two advanced memory features, memory advises and asynchronous prefetch, have been introduced. In this work, we evaluate the new features on two platforms that feature different CPUs, GPUs, and interconnects. We derive a benchmark suite for the experiments and stress the memory system to evaluate both in-memory and oversubscription performance. The results show that memory advises on the Intel-Volta/Pascal-PCIe platform bring negligible improvement for in-memory executions. However, when GPU memory is oversubscribed by about 50%, using memory advises results in up to 25% performance improvement compared to the basic CUDA Unified Memory. In contrast, the Power9-Volta-NVLink platform can substantially benefit from memory advises, achieving up to 34% performance gain for in-memory executions. However, when GPU memory is oversubscribed on this platform, using memory advises increases GPU page faults and results in considerable performance loss. The CUDA prefetch also shows different performance impact on the two platforms. It improves performance by up to 50% on the Intel-Volta/Pascal-PCI-E platform but brings little benefit to the Power9-Volta-NVLink platform.
Community shame:
Not yet rated
Community Contributions

Found the code? Know the venue? Think something is wrong? Let us know!

πŸ“œ Similar Papers

In the same crypt β€” Distributed Computing

Died the same way β€” πŸ‘» Ghosted