Compiler-Assisted Synthesis of Algorithm-Based Checking in Multiprocessors
IEEE Transactions on Computers
PVM: a framework for parallel distributed computing
Concurrency: Practice and Experience
Algorithm-Based Fault Tolerance on a Hypercube Multiprocessor
IEEE Transactions on Computers
A tutorial on Reed-Solomon coding for fault-tolerance in RAID-like systems
Software—Practice & Experience
IEEE Transactions on Parallel and Distributed Systems
The grid
Supercomputing '96 Proceedings of the 1996 ACM/IEEE conference on Supercomputing
A bandwidth latency tradeoff for broadcast and reduction
Information Processing Letters
Fault-tolerant matrix operations for parallel and distributed systems
Fault-tolerant matrix operations for parallel and distributed systems
Fault tolerant high performance computing by a coding approach
Proceedings of the tenth ACM SIGPLAN symposium on Principles and practice of parallel programming
Algorithm-Based Fault Tolerance for Matrix Operations
IEEE Transactions on Computers
HotDep'08 Proceedings of the Fourth conference on Hot topics in system dependability
High performance linpack benchmark: a fault tolerant implementation without checkpointing
Proceedings of the international conference on Supercomputing
Algorithm-based recovery for iterative methods without checkpointing
Proceedings of the 20th international symposium on High performance distributed computing
Tolerating correlated failures for generalized Cartesian distributions via bipartite matching
Proceedings of the 8th ACM International Conference on Computing Frontiers
Application-specific fault tolerance via data access characterization
Euro-Par'11 Proceedings of the 17th international conference on Parallel processing - Volume Part II
FTI: high performance fault tolerance interface for hybrid systems
Proceedings of 2011 International Conference for High Performance Computing, Networking, Storage and Analysis
Algorithm-based fault tolerance for dense matrix factorizations
Proceedings of the 17th ACM SIGPLAN symposium on Principles and Practice of Parallel Programming
Cooperative Application/OS DRAM fault recovery
Euro-Par'11 Proceedings of the 2011 international conference on Parallel Processing - Volume 2
Evaluating operating system vulnerability to memory errors
Proceedings of the 2nd International Workshop on Runtime and Operating Systems for Supercomputers
Convergence analysis of evolutionary algorithms in the presence of crash-faults and cheaters
Computers & Mathematics with Applications
Proceedings of the 18th ACM SIGPLAN symposium on Principles and practice of parallel programming
Evaluating the feasibility of using memory content similarity to improve system resilience
Proceedings of the 3rd International Workshop on Runtime and Operating Systems for Supercomputers
A study of application-level recovery methods for transient network faults
ScalA '13 Proceedings of the Workshop on Latest Advances in Scalable Algorithms for Large-Scale Systems
Future Generation Computer Systems
| Hi-index | 0.00 |
As the size of today's high performance computers increases from hundreds, to thousands, and even tens of thousands of processors, node failures in these computers are becoming frequent events. Although checkpoint/rollbaek-reovery is the typical technique to tolerate such failures, it often introduces a considerable overhead. Algorithm-based fault tolerance is a very cost-effective method to incorporate fault tolerance into matrix eomputations. However, previous algorithm-based fault tolerance methods for matrix computations are often derived using algorithms that are seldomly used in the practice of today's high performance matrix computations and have mostly focused on platforms where failed processors produce incorrect calculations. To fill this gap, this paper extends the existing algorithm-based fault tolerance to the volatile computing platform where the failied processor stops working and applies it to scalable high performance matrix computations with two dimensional block cyclic data distribution. We show the practicality of this technique by applying it to the ScaLAPACK/PBLAS matrix-matrix multiplication kernel. Experimental results demonstrate that the proposed approach is able to survive process failures with a very low performance overhead.