% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@ARTICLE{Tarraf:1035001,
author = {Tarraf, Ahmad and Schreiber, Martin and Cascajo, Alberto
and Besnard, Jean-Baptiste and Vef, Marc-André and Huber,
Dominik and Happ, Sonja and Brinkmann, André and Singh,
David E. and Hoppe, Hans-Christian and Miranda, Alberto and
Peña, Antonio J. and Machado, Rui and Garcia-Gasulla, Marta
and Schulz, Martin and Carpenter, Paul and Pickartz, Simon
and Rotaru, Tiberiu and Iserte, Sergio and Lopez, Victor and
Ejarque, Jorge and Sirwani, Heena and Carretero, Jesus and
Wolf, Felix},
title = {{M}alleability in {M}odern {HPC} {S}ystems: {C}urrent
{E}xperiences, {C}hallenges, and {F}uture {O}pportunities},
journal = {IEEE transactions on parallel and distributed systems},
volume = {35},
number = {9},
issn = {2161-9883},
address = {New York, NY},
publisher = {IEEE},
reportid = {FZJ-2025-00107},
pages = {1551 - 1564},
year = {2024},
abstract = {With the increase of complex scientific simulations driven
by workflows and heterogeneous workload profiles, managing
system resources effectively is essential for improving
performance and system throughput, especially due to trends
like heterogeneous HPC and deeply integrated systems with
on-chip accelerators. For optimal resource utilization,
dynamic resource allocation can improve productivity across
all system and application levels, by adapting the
applications’ configurations to the system's resources. In
this context, malleable jobs, which can change resources at
runtime, can increase the system throughput and resource
utilization while bringing various advantages for HPC users
(e.g., shorter waiting time). Malleability has received much
attention recently, even though it has been an active
research area for more than two decades. This article
presents the state-of-the-art of malleable implementations
in HPC systems, targeting mainly malleability in compute and
I/O resources. Based on our experiences, we state our
current concerns and list future opportunities for
research.},
cin = {JSC},
ddc = {004},
cid = {I:(DE-Juel1)JSC-20090406},
pnm = {5122 - Future Computing $\&$ Big Data Systems (POF4-512) /
DEEP-SEA - DEEP – SOFTWARE FOR EXASCALE ARCHITECTURES
(955606) / ADMIRE - Adaptive multi-tier intelligent data
manager for Exascale (956748) / TIME-X - TIME
parallelisation: for eXascale computing and beyond (955701)
/ Verbundprojekt: TIME-X - Parallelisierung zeitabhängiger
Simulationen für das zukünftige Supercomputing (16HPC047)
/ REGALE - An open architecture to equip next generation HPC
applications with exascale capabilities (956560)},
pid = {G:(DE-HGF)POF4-5122 / G:(EU-Grant)955606 /
G:(EU-Grant)956748 / G:(EU-Grant)955701 / G:(BMBF)16HPC047 /
G:(EU-Grant)956560},
typ = {PUB:(DE-HGF)16},
UT = {WOS:001272190100002},
doi = {10.1109/TPDS.2024.3406764},
url = {https://juser.fz-juelich.de/record/1035001},
}