Crac: Checkpoint-restart architecture for cuda with streams and uvm
T Jain, G Cooperman - SC20: International Conference for High …, 2020 - ieeexplore.ieee.org
The share of the top 500 supercomputers with NVIDIA GPUs is now over 25% and continues
to grow. While fault tolerance is a critical issue for supercomputing, there does not currently …
to grow. While fault tolerance is a critical issue for supercomputing, there does not currently …
Reinit: Evaluating the performance of global-restart recovery methods for mpi fault tolerance
Scaling supercomputers comes with an increase in failure rates due to the increasing
number of hardware components. In standard practice, applications are made resilient …
number of hardware components. In standard practice, applications are made resilient …
MANA for MPI: MPI-agnostic network-agnostic transparent checkpointing
Transparently checkpointing MPI for fault tolerance and load balancing is a long-standing
problem in HPC. The problem has been complicated by the need to provide checkpoint …
problem in HPC. The problem has been complicated by the need to provide checkpoint …
Shrink or substitute: handling process failures in HPC systems using in-situ recovery
RA Ashraf, S Hukerikar… - 2018 26th Euromicro …, 2018 - ieeexplore.ieee.org
Efficient utilization of today's high-performance computing (HPC) systems with complex
software and hardware components requires that the HPC applications are designed to …
software and hardware components requires that the HPC applications are designed to …
Software approaches for resilience of high performance computing systems: a survey
With the scaling up of high-performance computing systems in recent years, their reliability
has been descending continuously. Therefore, system resilience has been regarded as one …
has been descending continuously. Therefore, system resilience has been regarded as one …
LightPC: hardware and software co-design for energy-efficient full system persistence
We propose LightPC, a lightweight persistence-centric platform to make the system robust
against power loss. LightPC consists of hardware and software subsystems, each being …
against power loss. LightPC consists of hardware and software subsystems, each being …
Experimental findings on the sources of detected unrecoverable errors in gpus
FF dos Santos, S Malde, C Cazzaniga… - … on Nuclear Science, 2022 - ieeexplore.ieee.org
We investigate the sources of detected unrecoverable errors (DUEs) in graphics processing
units (GPUs) exposed to a neutron beam. Illegal memory accesses and interface errors are …
units (GPUs) exposed to a neutron beam. Illegal memory accesses and interface errors are …
Towards Efficient Cache Allocation for High-Frequency Checkpointing
While many HPC applications are known to have long runtimes, this is not always because
of single large runs: in many cases, this is due to ensembles composed of many short runs …
of single large runs: in many cases, this is due to ensembles composed of many short runs …
Checkpointing strategies to tolerate non-memoryless failures on HPC platforms
This article studies checkpointing strategies for parallel applications subject to failures. The
optimal strategy to minimize total execution time, or makespan, is well known when failure …
optimal strategy to minimize total execution time, or makespan, is well known when failure …
Job migration in hpc clusters by means of checkpoint/restart
M Rodríguez-Pascual, J Cao, JA Moríñigo… - The Journal of …, 2019 - Springer
Until now, jobs running on HPC clusters were tied to the node where their execution started.
We have removed that limitation by integrating a user-level checkpoint/restart library into a …
We have removed that limitation by integrating a user-level checkpoint/restart library into a …