Benchmarking the MIT LL HPCMP DHPI System

The High Performance Computing Modernization Program (HPCMP) obtains new supercomputing hardware each year through a rigorous acquisition process known as technology insertion (TI). TI is based primarily on application benchmarks. The suite of applications used spans many disciplines and is carefully constructed to accurately reflect the high Performance Computing Modernization Program (HPCMP) computational workload. The application codes GAMESS, OOCORE, and RF-CTH were used to represent the contributions to the workload in the areas of computational chemistry and materials science (CCM), computational electromagnetics and acoustics (CEA), and computational structural mechanics (CSM), respectively. This paper will compare the performance of the application test cases GAMESS standard/large, OOCORE standard/large, and RF-CTH standard/large on major HPCMP high performance computing (HPC) architectures.

Each year the High Performance Computing Modernization Program (HPCMP) obtains new supercomputing hardware through a rigorous acquisition process known as technology insertion (TI). TI is based primarily on application benchmarks. Its suite of applications spans many disciplines and is carefully constructed to ensure that it accurately reflects the HPCMP workload. For the most recent TI, (TI-05), the application codes AERO, AVUS, and OVERFLOW 2 were used to represent the computational fluid dynamics (CFD) portion of the workload. This paper will compare the performance of the major HPCMP architectures in executing each of the following CFD application test cases found within the TI-05 benchmarking suite: AERO standard, AVUS standard/large, and OVERFLOW 2 standard/large.

The US Department of Defense(DoD), relies on high-performance computing as a key enabling technology. The breadth of application is impressive, impacting a wide range of activities from basic research to weapons testing. This edition of Computing in Science & Engineering highlights a subset of HPC uses within the DoD, specifically those supported by the DoD High-Performance Computing Modernization Program (HPCMP).

The Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) code is part of the Department of Defense High Performance Computing Modernization Program (HPCMP) technology insertion (TI) benchmarking suite of applications. As a component of the TI benchmarking applications, LAMMPS is a significant contributor to Computational Chemistry and Materials Science requirements within the DoD. Ensuring its optimal performance on HPCMP resources is a high priority. The CCM-KY5-003 User Productivity Enhancements and Technology Transfer (PET) project was created to profile and optimize LAMMPS to improve its performance and efficiency on two HPCMP assets, Eagle (Altix 3700) and JVN(Linux Cluster). Profiling efforts were completed on Eagle using the Tuning and Analysis Utilities (TAU) application in conjunction with the Performance Application Programming Interface (PAPI). The time and hardware counter data was analyzed with ParaProf and PerfExplorer, while the trace data was automatically analyzed by the Kit for Objective Judgment and Knowledge-based Detection of Performance Bottlenecks (KOJAK). The profiling effort used three different model systems, where each uses a different physical potential. Several opportunities for performance improvement were identified in the critical portions of the code, such as loop-static branching conditions, multi-dimensional C-arrays and redundant integer, and floating-point operations. In addition to algorithmic changes, compiler options and machine-specific instructions were also investigated. The performance of the code increased from 1.3x to 3.5x on Eagle and 1.2x to 1.8x on JVN depending on the physical potential employed

The parallel WASH123D is designed to simulate watershed systems using a coupled system of one-dimensional (1-D) channels, 2-D overland areas, and 3- D subsurface media on parallel scalable computers. The U.S. Department of Defense High Performance Computing Modernization Program funds the parallelization of the watershed model through the Common High Performance Computing Software Initiative in order to solve one aspect of the battlespace environment, which in- cludes space, weather, ocean, and soil, to develop a complete coupled battlespace environment. Tasks in this project include parallel algorithm development, software toolkit development, and performance studies. Currently, the parallel version, which includes coupled 2-D overland and 3-D subsurface flows, is completed and packaged for application projects, such as calibration and validation of the coupled 2-D/3-D Biscayne Bay Coastal Wetland inland flow model presented in this paper, on various computer architectures.

HPC Benchmark Assessment with Statistical Analysis

This manuscript is a copy of a successful proposal to the High Performance Computing Modernization Program of the Department of Defense for a Grand Challenge Award. Although it does not relate directly to high temperature ceramics (the topic of this conference proceedings), the ideas contained herein are directly pertinent to the realistic simulation of ceramics. This grand challenge project will predict the dynamic and temperature dependent behavior of Micro Electro Mechanical Systems (MEMS) devices from their atomic constituents by utilizing recently developed parallel codes which enable a coupling of length scales. This powerful methodology will be applied to two MEMS systems thereby providing a new paradigm in device design and prediction and in the relevance of atomistic simulation.

Purpose: High-Performance Computing (HPC) is the technique of increasing the overall processing capacity of the system, and achieving high performance in solving large problems in science, engineering, or business. This research paper investigated the High-Performance Computing industry using exploratory research methodology and a listing of ABCD analysis is done. Along with this, the identification of various applications and grand challenges in HPC and various Software Tools required for industry-specific application areas of HPC is done. Design/Methodology/Approach: Qualitative exploratory research methodology to perform Industry Analysis of High-Performance Computing using ABCD Analysis. This research paper is written with the literature review of High-Performance Computing (HPC) industries and various software tools available to use in those industries. Findings/Results: Finding of this research paper is the identification of various grand challenges in HPC industries like healthcare and medical field, engineering and design, aerospace sector, urban planning, and infrastructure, finance, and business. It also lists different HPC software tools that can be used to solve these challenges, along with the findings of ABCD analysis of HPC industry. Originality/Value: Listing of ABCD Analysis (Advantages, Benefits, Constraints, and Disadvantages) of High-Performance Computing industry. Paper Type: Industry Analysis research paper

The lightweight nature, application portability, and deployment flexibility of containers is driving their widespread adoption in cloud solutions. Data analysis and deep learning (DL)/machine learning (ML) applications have especially benefited from containerization. As such data analysis is adopted in high performance computing (HPC), the need for container support in HPC has become paramount. However, containers face crucial performance and I/O challenges in HPC. One obstacle is that while there have been HPC containers, such solutions have not been thoroughly investigated, especially from the aspect of their impact on the crucial HPC I/O throughput. To this end, this paper provides a first-of-its-kind empirical analysis of state-of-the-art representative container solutions (Docker, Podman, Singularity, and Charliecloud) in HPC environments. We also explore how containers interact with an HPC parallel file system like Lustre. We present the design of an analysis framework that is deployed on all nodes in an HPC environment, and captures CPU, memory, network, and file I/O statistics from the nodes and the storage system. We are able to garner key insights from our analysis, e.g., Charliecloud outperforms other container solutions in terms of container start-up time, while Singularity and Charliecloud are equivalent in I/O throughput. But this comes at a cost, as Charliecloud invokes the most metadata and I/O operations on the underlying Lustre file system. By identifying such trade-offs and optimization opportunities, we can enhance HPC containers performance and the ML/DL applications that increasingly rely on them.

Density Functional Theory calculations with plane waves and pseudopotentials represent one of the most important simulation techniques in high performance computing. Together with parallel linear algebra (ZGEMM and matrix diagonalization), the most important bottleneck results from the Fast Fourier Transform (FFT), required, for example, when the local potential is applied to the wavefunction. In these calculations, the existence of a cutoff on the plane waves is reflected on a spherical domain for the FFT. After a 1D FFT is performed on pencils distributed among processors, data is transposed with a MPI_Alltoall and a 2D FFT is executed [2]. Typically, the workload of the FFT is not particularly high, since grid sizes do not exceed (103 102)3. However, the load distribution is crucial and the consequent impact of collective communications becomes a critical factor for achieving a high parallel efficiency. Quantum ESPRESSO [3] is one of the most used codes based on plane-wave DFT in the community of material science. It has been successfully ported and optimized on a large number of HPC infrastructures all over the world. The parallel structure of Quantum ESPRESSO is mainly based on several layers of MPI communicators, plus a finer grain OpenMP parallelization. Recently, the parallelization structure of the FFT was deeply refactored. The combination of two different data distributions, i.e. bands and taskgroups, allow the underlyinghardware to be hierarchically filled and two different layers of communications to be tuned. In particular, with sufficient memory, by tuning the number of taskgroups one can fit all the data required to perform a single 3D FFT reducing the impact of the MPI_Alltoall between the 1D and 2D FFTs. In order to better check the results of the parametrization of the parallel distributions, a miniapp [1] containing only the FFT kernel was extracted from the Quantum ESPRESSO distribution. This miniapp is also important for the future activity of code design of novel architectures. We present and discuss the profiling data obtained from the QE-FFT miniapp and the impact on the communication pattern deriving from the choice of the parallelization parameters.

High Performance Computing (HPC) is used to achieve best of performance. However, we see that power and energy are both critical concerns in High performance computing applications. One of the biggest challenges in HPC is power efficiency. Power and energy consumption has to be minimized to reduce the cost without much change in the performance of the application. One basic requirement to achieve this power efficiency is to measure the power consumption of application runs. If we cannot identify the power profile of an application, improvising the power utilization becomes more challenging. Although, external electronic watt meters exist to perform the analysis, software techniques to detect the power usage have been found to be more convenient for dynamic runs of the application. Also, the use of external hardware implies extra costs. Additionally, power has to be supplied to the device for its operation. Previous efforts to develop such a system have involved the use of expensive and bulky hardware technology and software applications that are not very accurate. We propose a software tool named PHOENIX that measures the power consumption and identifies the power profiles of the applications. The results are validated using a hardware power meter. We use rodinia benchmark suite to verify our results. Our results show that we achieve an average of 88% accuracy when compared to hardware power meter.

The Naval Oceanographic Office (NAVOCEANO) Department of Defense (DoD) Major Shared Resource Center (MSRC) operates as a DoD shared high performance computing (HPC) center, serving over 2000 users nationwide. Scientific visualization is an essential element of HPC that provides the methodology to explore, define, and present the results of computations. Accurate and understandable representation of data impacts a project in several ways: it helps scientists understand the physical phenomena they are studying, it helps communicate this work to their colleagues, and it helps explain the work and its significance to the public. Our scientific visualization team, in conjunction with researchers at our facility and other DoD facilities, have accomplished numerous visualizations of large data sets. In this paper we present examples of the methods, techniques, and results of our efforts to visualize observational data and the output of numeric ocean models.

The Fast Fourier Transform (FFT) is of primary importance and a fundamental kernel in many computationally intensive scientific applications. In this paper we investigate its performance on the Sony-Toshiba-IBM Cell Broadband Engine, a heterogeneous multicore chip architected for intensive gaming applications and high performance computing. The Cell processor consists of a traditional microprocessor (called the PPE) that controls eight SIMD co-processing units called synergistic processor elements (SPEs). We exploit the architectural features of the Cell processor to design an efficient parallel implementation of Fast Fourier Transform (FFT). While there have been several attempts to develop a fast implementation of FFT on the Cell, none have been able to achieve high performance for input series with several thousand complex points. We use an iterative out-of-place approach to design our parallel implementation of FFT with 1K to 16K complex input samples and attain a single precision performance of 18.6 GFLOP/s on the Cell. Our implementation beats FFTW on Cell by several GFLOP/s for these input sizes and outperforms Intel Duo Core (Woodcrest) for inputs of greater than 2K samples. To our knowledge we have the fastest FFT for this range of complex inputs.

The suitability of next-generation high-performance computing systems for petascale simulations will depend on various performance factors attributable to processor, memory, local and global network, and input/output characteristics. In this paper, we evaluate performance of new dual-core SGI Altix 4700, quad-core SGI Altix ICE 8200, and dual-core IBM POWER5+ systems. To measure performance, we used micro-benchmarks from High Performance Computing Challenge (HPCC), NAS Parallel Benchmarks (NPB), and four real-world applications---three from computational fluid dynamics (CFD) and one from climate modeling. We used the micro-benchmarks to develop a controlled understanding of individual system components, then analyzed and interpreted performance of the NPBs and applications. We also explored the hybrid programming model (MPI+OpenMP) using multi-zone NPBs and the CFD application OVERFLOW-2. Achievable application performance is compared across the systems. For the ICE platform, we also investigated the effect of memory bandwidth on performance by testing 1, 2, 4, and 8 cores per node.

High-Performance Computing, HPC, has become one of the most active computer science fields. Driven mainly by the need for high processing capabilities required by algorithms from many areas, such as Big Data, Artificial Intelligence, Data Science, and subjects related to chemistry, physics, and biology, the state-of-art algorithms from these fields are notoriously demanding computer resources. Therefore, choosing the right computer system to optimize their performance is paramount. This article presents the main challenges of future supercomputer systems, highlighting the areas that demand the most of HPC servers; the new architectures, including heterogeneous processors composed of artificial intelligence chips, quantum processors, the adoption of HPC on cloud servers; and the challenges of software developers when facing parallelizing applications. We also discuss challenges regarding non-functional requirements, such as energy consumption and resilience.