Investigating the effect of varying block size on power and energy consumption of GPU kernels

被引：4

作者：

Ikram, Muhammad Jawad ^{[1
]}

Saleh, Mostafa Elsayed ^{[2
]}

Al-Hashimi, Muhammad Abdulhamid ^{[2
]}

Abulnaja, Osama Ahmed ^{[2
]}

机构：

[1] Jeddah Int Coll, Jeddah 23831, Saudi Arabia

[2] King Abdulaziz Univ, Jeddah 21589, Saudi Arabia

来源：

JOURNAL OF SUPERCOMPUTING | 2022年 / 78卷 / 13期

关键词：

Block size; Exascale computing; GPU; Power and energy; EFFICIENCY; DESIGN;

D O I：

10.1007/s11227-022-04473-9

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

Power consumption is likely to remain a significant concern for exascale performance in the foreseeable future. In addition, graphics processing units (GPUs) have become an accepted architectural feature for exascale computing due to their scalable performance and power efficiency. In a recent study, we found that we can achieve a reasonable amount of power and energy savings based on the selection of algorithms. In this research, we suggest that we can save more power and energy by varying the block size in the kernel configuration. We show that we may attain more savings by selecting the optimum block size while executing the workload. We investigated two kernels on NVIDIA Tesla K40 GPU, a Bitonic Mergesort and Vector Addition kernels, to study the effect of varying block sizes on GPU power and energy consumption. The study should offer insights for upcoming exascale systems in terms of power and energy efficiency.

引用

页码：14919 / 14939

页数：21

共 50 条

[1] Investigating the effect of varying block size on power and energy consumption of GPU kernels
Muhammad Jawad Ikram
Mostafa Elsayed Saleh
Muhammad Abdulhamid Al-Hashimi
Osama Ahmed Abulnaja
The Journal of Supercomputing, 2022, 78 : 14919 - 14939
[2] Energy consumption of CUDA kernels with varying thread topology
Dressler, Sebastian
Steinke, Thomas
COMPUTER SCIENCE-RESEARCH AND DEVELOPMENT, 2014, 29 (02): : 113 - 121
[3] Statistical Power and Energy Modeling of Multi-GPU kernels
Ghosh, Sayan
Chandrasekaran, Sunita
Chapman, Barbara M.
2012 SC COMPANION: HIGH PERFORMANCE COMPUTING, NETWORKING, STORAGE AND ANALYSIS (SCC), 2012, : 1516 - 1516
[4] A Simple Model for Portable and Fast Prediction of Execution Time and Power Consumption of GPU Kernels
Braun, Lorenz
Nikas, Sotirios
Song, Chen
Heuveline, Vincent
Froening, Holger
ACM TRANSACTIONS ON ARCHITECTURE AND CODE OPTIMIZATION, 2021, 18 (01)
[5] Statistical Modeling of Power/Energy of Scientific Kernels on a Multi-GPU system
Ghosh, Sayan
Chandrasekaran, Sunita
Chapman, Barbara
2013 INTERNATIONAL GREEN COMPUTING CONFERENCE (IGCC), 2013,
[6] Investigating the effect of design patterns on energy consumption
Feitosa, Daniel
Alders, Rutger
Ampatzoglou, Apostolos
Avgeriou, Paris
Nakagawa, Elisa Yumi
JOURNAL OF SOFTWARE-EVOLUTION AND PROCESS, 2017, 29 (02)
[7] EFFECT OF VARYING SIZE OF SCLERAL FLAP AND CORNEAL BLOCK ON TRABECULECTOMY
STARITA, RJ
FELLMAN, RL
SPAETH, GL
PORYZEES, EM
OPHTHALMIC SURGERY AND LASERS, 1984, 15 (06): : 484 - 487
[8] MIPP: A Microbenchmark Suite for Performance, Power, and Energy Consumption Characterization of GPU architectures
Bombieri, Nicola
Busato, Federico
Fummi, Franco
Scala, Michele
2016 11TH IEEE INTERNATIONAL SYMPOSIUM ON INDUSTRIAL EMBEDDED SYSTEMS (SIES), 2016,
[9] Analyzing the effect of a block FEC algorithm's symbol size on energy consumption in wireless sensor networks
Ahn, Jong-Suk
Lee, Young-Su
Yoon, Jong-Hyuk
Lee, Kang-Woo
UBIQUITOUS COMPUTING SYSTEMS, PROCEEDINGS, 2006, 4239 : 440 - 453
[10] Impacts of optimization strategies on performance, power/energy consumption of a GPU based parallel reduction
Phuong Thi Yen
Lee Deok-Young
Lee Jeong-Gun
Journal of Central South University, 2017, 24 (11) : 2624 - 2637

← 1 2 3 4 5 →