Out-of-core simulation systems produce and/or consume a massive amount of data that cannot fit on a single compute node memory and that usually needs to be read and/or written back and forth during computation. I/O data movement may thus represent a bottleneck in large-scale simulations. To increase I/O bandwidth, high-end supercomputers are equipped with hierarchical storage subsystems such as node-local and remote-shared NVMe and SSD-based Burst Buffers. Advanced caching systems have recently been developed to efficiently utilize the multi-layered nature of the new storage hierarchy. Utilization of software components results in more efficient data accesses, at the cost of reduced computation kernel performance and limited numbers of simultaneous applications that can utilize the additional storage layers. We introduce MultiLayered Buffer Storage (MLBS), a data object container that provides novel methods for caching and prefetching data in out-of-core scientific applications to perform asynchronously expensive I/O operations on systems equipped with hierarchical storage. The main idea consists in decoupling I/O operations from computational phases using dedicated hardware resources to perform expensive context switches. MLBS monitors I/O traffic in each storage layer allowing fair utilization of shared resources while controlling the impact on kernels' performance. By continually prefetching up and down across all hardware layers of the memory/storage subsystems, MLBS transforms the original I/O-bound behavior of evaluated applications and shifts it closer to a memory-bound regime. Our evaluation on a Cray XC40 system for a representative I/O-bound application, seismic inversion, shows that MLBS outperforms state-of-the-art filesystems, i.e., Lustre, Data Elevator and DataWarp by 6.06X, 2.23X, and 1.90X, respectively.

This paper presents APlug, a framework for automatic tuning of large scale applications of many independent tasks. APlug suggests the best decomposition of the original computation into smaller tasks and the best number of CPUs to use, in order to meet user-specific constraints. We show that the problem is not trivial because there is large variability in the execution time of tasks, and it is possible for a task to occupy a CPU by performing useless computations. APlug collects a sample of task execution times and builds a model, which is then used by a discrete event simulator to calculate the optimal parameters. We provide a C++ API and a stand-alone implementation of APlug, and we integrate it with three typical applications from computational chemistry, bioinformatics, and data mining. A scenario for optimizing resources utilization is used to demonstrate our framework. We run experiments on 16,384 CPUs on a supercomputer, 480 cores on a Linux cluster and 80 cores on Amazon EC2, and show that APlug is very accurate with minimal overhead.