A New and Efficient Algorithm-Based Fault Tolerance Scheme for A Million Way Parallelism

TL;DR

This paper introduces a novel, efficient algorithm-based fault tolerance scheme for large-scale HPC systems, capable of handling failures during execution without halting, suitable for systems reaching Exaflops scale.

Contribution

The paper proposes a new fault tolerance method that replaces corrupted data with redundancy during execution and includes a background recovery process, effective at Exaflops scale.

Findings

Effective even at Exaflops scale

Reduces fault tolerance overhead

Validated on SiCortex SC5832

Abstract

Fault tolerance overhead of high performance computing (HPC) applications is becoming critical to the efficient utilization of HPC systems at large scale. HPC applications typically tolerate fail-stop failures by checkpointing. Another promising method is in the algorithm level, called algorithmic recovery. These two methods can achieve high efficiency when the system scale is not very large, but will both lose their effectiveness when systems approach the scale of Exaflops, where the number of processors including in system is expected to achieve one million. This paper develops a new and efficient algorithm-based fault tolerance scheme for HPC applications. When failure occurs during the execution, we do not stop to wait for the recovery of corrupted data, but replace them with the corresponding redundant data and continue the execution. A background accelerated recovery method is also proposed to rebuild redundancy to tolerate multiple times of failures during the execution. To demonstrate the feasibility of our new scheme, we have incorporated it to the High Performance Linpack. Theoretical analysis demonstrates that our new fault tolerance scheme can still be effective even when the system scale achieves the Exaflops. Experiment using SiCortex SC5832 verifies the feasibility of the scheme, and indicates that the advantage of our scheme can be observable even in a small scale.