Efficient checkpoint/Restart of CUDA applications

被引：0

作者：

Nukada, Akira ^{[1
]}

Suzuki, Taichiro ^{[3
]}

Matsuoka, Satoshi ^{[2
,3
]}

机构：

[1] Univ Tsukuba, CCS, 1-1-1 Tennodai, Tsukuba, Ibaraki 3058577, Japan

[2] RIKEN R CCS, 7-1-26 Minatojima Minami Machi,Chuo Ku, Kobe, Hyogo 6500047, Japan

[3] Tokyo Inst Technol, 2-12-1 Oookayama,Meguro Ku, Tokyo 1528550, Japan

来源：

PARALLEL COMPUTING | 2023年 / 116卷

关键词：

Checkpoint and restart; NVIDIA CUDA; GPU;

D O I：

10.1016/j.parco.2023.103018

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

We present NVCR which enables transparent checkpoint and restart of CUDA applications. NVCR, works as an extension of major system-level checkpoint software such as BLCR and DMTCP, employs proxy-process and application accesses GPU devices via the proxy-process to improve the compatibility with latest CUDA runtime software. To reduce the overhead of inter-process communications, NVCR efficiently uses SYSV IPC shared memory as CUDA pinned memory. Performance evaluations using micro benchmarks and Amber as a real application show that NVCR' overhead is acceptably low.

引用

页数：9

共 50 条

[1] CheCUDA: A Checkpoint/Restart Tool for CUDA Applications
Takizawa, Hiroyuki
Sato, Katsuto
Komatsu, Kazuhiko
Kobayashi, Hiroaki
2009 INTERNATIONAL CONFERENCE ON PARALLEL AND DISTRIBUTED COMPUTING, APPLICATIONS AND TECHNOLOGIES (PDCAT 2009), 2009, : 408 - +
[2] A Checkpoint/Restart Scheme for CUDA Applications with Complex Memory Hierarchy
Zhang, Yulu
Guo, Xinyuan
Jiang, Hai
Li, Kuan-Ching
2013 14TH ACIS INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING, ARTIFICIAL INTELLIGENCE, NETWORKING AND PARALLEL/DISTRIBUTED COMPUTING (SNPD 2013), 2013, : 247 - 252
[3] Cricket: A virtualization layer for distributed execution of CUDA applications with checkpoint/restart support
Eiling, Niklas
Baude, Jonas
Lankes, Stefan
Monti, Antonello
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE, 2022, 34 (14):
[4] A Checkpoint/Restart Scheme for CUDA Programs with Complex Computation States
Jiang, Hai
Zhang, Yulu
Jenness, Jeff
Li, Kuan-Ching
INTERNATIONAL JOURNAL OF NETWORKED AND DISTRIBUTED COMPUTING, 2013, 1 (04) : 196 - 212
[5] A checkpoint/restart scheme for CUDA programs with complex computation states
Jiang H.
Zhang Y.
Jenness J.
Li K.-C.
International Journal of Networked and Distributed Computing, 2013, 1 (4) : 196 - 212
[6] CRAC: Checkpoint-Restart Architecture for CUDA with Streams and UVM
Jain, Twinkle
Cooperman, Gene
PROCEEDINGS OF SC20: THE INTERNATIONAL CONFERENCE FOR HIGH PERFORMANCE COMPUTING, NETWORKING, STORAGE AND ANALYSIS (SC20), 2020,
[7] CRUM: Checkpoint-Restart Support for CUDA's Unified Memory
Garg, Rohan
Mohan, Apoorve
Sullivan, Michael
Cooperman, Gene
2018 IEEE INTERNATIONAL CONFERENCE ON CLUSTER COMPUTING (CLUSTER), 2018, : 302 - 313
[8] Checkpoint Restart Support for Heterogeneous HPC Applications
Parasyris, Konstantinos
Keller, Kai
Bautista-Gomez, Leonardo
Unsal, Osman
2020 20TH IEEE/ACM INTERNATIONAL SYMPOSIUM ON CLUSTER, CLOUD AND INTERNET COMPUTING (CCGRID 2020), 2020, : 242 - 251
[9] Efficient Encoding and Reconstruction of HPC Datasets for Checkpoint/Restart
Zhang, Jialing
Zhuo, Xiaoyan
Moon, Aekyeung
Liu, Hang
Son, Seung Woo
2019 35TH SYMPOSIUM ON MASS STORAGE SYSTEMS AND TECHNOLOGIES (MSST 2019), 2019, : 79 - 91
[10] cudaCR: An In-kernel Application-level Checkpoint/Restart Scheme for CUDA-enabled GPUs
Pourghassemi, Behnam
Chandramowlishwaran, Aparna
2017 IEEE INTERNATIONAL CONFERENCE ON CLUSTER COMPUTING (CLUSTER), 2017, : 725 - 732

← 1 2 3 4 5 →