insideHPC

MemVerge and Open Source Community Partnering to Protect Distributed HPC Apps with DMTCP

ST. LOUIS – NOVEMBER 15, 2021 – At SC21 today, MemVerge and the DMTCP Project announced a partnership designed to accelerate development and adoption of long-awaited Distributed MultiThreaded ...
The Next Platform

Memory Snapshots Bring Checkpointing Into The 21st Century

When you have a massively distributed computing job that can take months to run across thousands to hundreds of thousands of compute elements, one software hardware or software crash can mean losing ...
insideHPC

Checkpointing the Un-checkpointable: MANA and the Split-Process Approach

In this video from the MVAPICH User Group, Gene Cooperman from Northeastern University presents: Checkpointing the Un-checkpointable: MANA and the Split-Process Approach. Checkpointing is the ability ...
Rochester Institute of Technology

Rafique and Maurya receive Best Paper Award at High Performance Distributed Computing Conference

M. Mustafa Rafique, associate professor of computer science, and Avinash Maurya, a computer science Ph.D. student, received the Best Paper Award from the Association for Computing Machinery ...
EurekAlert!

A fault tolerant framework for distributed training with negligible overhead

In the context of deep learning model training, checkpoint-based error recovery techniques are a simple and effective form of fault tolerance. By regularly saving the ...
Nature

Fault Tolerance in High-Performance Computing Systems

The ever-growing scale of high-performance computing systems, particularly with the transition to exascale computing, has underscored the critical need for robust fault tolerance. As these systems ...