Proceedings of the 2011 ACM International Workshop on Many Task Computing on Grids and Supercomputers 2011
DOI: 10.1145/2132876.2132887
|View full text |Cite
|
Sign up to set email alerts
|

Towards scalable I/O architecture for exascale systems

Abstract: High performance computing (HPC) has crossed the Petaflop mark and is reaching the Exaflop range quickly. The exascale system is projected to have millions of nodes, with thousands of cores for each node. At such an extreme scale, the substantial amount of concurrency can cause a critical contention issue for I/O system. The contention can destroy the request locality, increase the access latency, and waste the precious I/O interconnection bandwidth. This study proposes a dynamically coordinated I/O architectu… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

0
3
0
1

Year Published

2012
2012
2019
2019

Publication Types

Select...
4
2

Relationship

0
6

Authors

Journals

citations
Cited by 7 publications
(4 citation statements)
references
References 27 publications
0
3
0
1
Order By: Relevance
“…systems with Exaflop peak compute capacity. When reaching this level of capacity the number of cores per node is expected to increase significantly [6] while node memory capacity and I/O latency and bandwidth are expected to not be able grow at the same pace (due to power consumption and technology limitations respectively). Thus, the gap between CPU and I/O capacity is expected to increase, and I/O to become a performance cap for certain applications.…”
Section: Hpc Trendsmentioning
confidence: 99%
See 2 more Smart Citations
“…systems with Exaflop peak compute capacity. When reaching this level of capacity the number of cores per node is expected to increase significantly [6] while node memory capacity and I/O latency and bandwidth are expected to not be able grow at the same pace (due to power consumption and technology limitations respectively). Thus, the gap between CPU and I/O capacity is expected to increase, and I/O to become a performance cap for certain applications.…”
Section: Hpc Trendsmentioning
confidence: 99%
“…• I/O dedicated nodes present a similar function to the burst buffers but out of the compute nodes, higher capacity and slightly larger latency [6].…”
Section: Hpc Trendsmentioning
confidence: 99%
See 1 more Smart Citation
“…Разработанные аппаратно-программные средства ввода/вывода данных и проведенные исследования актуальны для разрабатываемых новых суперкомпьютерных систем эксафлопсного класса [25,26].…”
Section: заключениеunclassified