Search

With the computing power of High-Performance Computing (HPC) systems having stepped into the exascale era, more complex problems can be solved with scientific applications on a large scale. However, due to the significant performance gap between computing nodes and storage subsystems, suboptimal design for the Input/Output (I/O) module will significantly impede the efficiency of scientific applications, especially for the ubiquitous atmosphere applications. Two-phase I/O implemented in N-to-1 mode creates a serious bottleneck that hinders the scalability for the Model for Prediction Across Scales-Atmosphere (MPAS-A) on the new generation Sunway supercomputer. To address the I/O problem, we apply a custom data reorganization method to enable N-to-M I/O mode to exploit the parallel file system's performance and limit the data transfer among MPI ranks to a restricted scope to alleviate communication overhead. Moreover, we have conducted several methods to accelerate the computations, including the redesign for tracer transport, a hybrid buffering scheme, and a three-level parallelization scheme, which allows MPAS-A to use all heterogeneous computing resources efficiently. Experimental results show admirable scalability and efficiency of our I/O method, which achieves speedups of 41× and 58.9× for input and output compared with the raw I/O method on 30,000 MPI ranks. By scaling MPAS-A to 39 million heterogeneous cores, we demonstrate the necessity of a well-constructed I/O module for a real-world atmosphere application. Speed tests show that our optimization methods obtain good results for computations, and MPAS-A achieves a speed of 0.82 Simulated Day per Hour (SDPH) and 0.76 parallel efficiency of strong scaling with 600,000 MPI ranks. [ABSTRACT FROM AUTHOR]

Distributed-memory graph algorithms are fundamental enablers in scientific computing and analytics workflows. A majority of graph algorithms rely on the graph neighborhood communication pattern, i.e., repeated asynchronous communication between a vertex and its neighbors in the graph. The pattern is adversarial for communication software and hardware due to high message injection rates and input-dependent, many-to-one traffic with variable destinations and volumes. We present benchmarks and performance analysis of graph neighborhood communication on modern large-scale network interconnects from four supercomputers: ALCF Theta, NERSC Cori, OLCF Summit and R-CCS Fugaku. Our benchmarks characterize communication from the perspectives of latency and throughput. Benchmark parameters make it possible to mimic the behaviors of complex applications on real world or synthetic graphs by varying work distribution, remote edges, message volume, and per-vertex work. We find that minor changes in the input graph can substantially increase latencies; and contention can develop in memory caches and network stacks before contention in the network itself. Further, latencies and contention vary significantly for different graph neighborhoods, motivating the need for exploring asynchronous algorithms in greater detail. When adding work, load imbalance on real-world graphs can be pronounced: latencies for the 99th percentile were 8–128× than the corresponding average latencies. Our results help analysts and developers understand the performance implications of this important pattern, especially for the impending exascale platforms. [ABSTRACT FROM AUTHOR]

In this special section we bring you a practice and experience effort in reproducibility for large-scale computational science at SC20. This section includes nine critiques, each by a student team that reproduced results from a paper published at SC19, during the following year’s Student Cluster Competition. The paper is also included in this section and has been expanded upon, now including an analysis of the outcomes of the students’ reproducibility experiments. Lastly, this special section encapsulates a variety of advances in reproducibility in the SC conference series technical program. [ABSTRACT FROM AUTHOR]