Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “NERSC”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

HEPCloud Operations at Fermilab—The First Five Years

The HEPCloud Facility at Fermilab has now been in production operation for five years. This facility is a unified provisioning gateway to US high performance computing centers, including NERSC, OLCF, and ALCF, other large supercomputers run by the NSF, and commercial clouds. HEPCloud delivers hundreds of millions of core-hours yearly for CMS. HEPCloud also serves other Fermilab experiments including DUNE, Mu2e, Muon g-2, and NOvA. In this paper we present the practical considerations of operating a distributed facility such as HEPCloud. We also mention some of the interesting research and development that HEPCloud has been used for including GPU-based machine learning inference servers, and tests of Quantum Computing.

Timm, Steven [Fermilab]↗
DOE OSTI2023

Planting a Lyman alpha forest on ABACUSSUMMIT

The full-shape correlations of the Lyman alpha (Ly$α$) forest contain a wealth of cosmological information through the Alcock Paczynski effect. However, these measurements are challenging to model without robustly testing and verifying the theoretical framework used for analyzing them. Here, we leverage the accuracy and volume of the $N$-body simulation suite ABACUSSUMMIT to generate high-resolution Ly$α$ skewers and quasi-stellar object (QSO) catalogs. One of the main goals of our mocks is to aid in the full-shape Ly$α$ analysis planned by the Dark Energy Spectroscopic Instrument (DESI) team. We provide optical depth skewers for six of the fiducial cosmology base-resolution simulations ($L$ box = 2 $h$ -1 Gpc, $N$ = 6912 3 ) at $z$ = 2.5. We adopt a simple recipe based on the Fluctuating Gunn-Peterson Approximation (FGPA) for constructing these skewers from the matter density in an $N$-body simulation and calibrate it against the 1D and 3D Ly$α$ power spectra extracted from the hydrodynamical simulation IllustrisTNG (TNG; $L$ box = 205 $h$ -1 Mpc, $N$ = 2500 3 ). As an important application, we study the non-linear broadening of the baryon acoustic oscillation (BAO) peak and show the cross-correlation between DESI-like QSOs and our Ly$α$ forest skewers. We find differences on small scales between the Kaiser approximation prediction and our mock measurements of the Ly$α$×QSO cross-correlation, which would be important to account for in upcoming analyses. The ABACUSSUMMIT Ly$α$ forest mocks open up the possibility for improved modelling of cross correlations between Ly$α$ and cosmic microwave background (CMB) lensing and Ly$α$ and QSOs, and for forecasts of the 3-point Ly$α$ correlation function. Furthermore, our catalogues and skewers are publicly available on Globus via the National Energy Research Scientific Computing Center (NERSC) (full link under Data Availability).

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

Containers for Massive Ensemble of I/O Bound Hierarchical Coupled Simulations

We present our experience using containers to scale up a massive ensemble of coupled I/O bound workloads on the NERSC Cori supercomputer. We describe the design of a hierarchical simulation structure using the Integrated Plasma Simulator (IPS) that enables the flexible execution of coupled simulations at the system, node, and core level using the same coupling abstraction and API. The hierarchical design allows for the node-level execution to be efficiently executed using containers while not impacting the structure of the simulation at the system level. We demonstrate the viability of the approach by presenting experimental results from applications in coupled fusion plasma simulations that illustrate the performance impact of using containers to deploy the node-level workloads, in conjunction with the user mountable XFS file systems to ameliorate the load on the Lustre parallel file system. We also present results from production runs showing the ability of the ensemble simulations to scale to hundreds of Cori Haswell nodes, with little or no overhead.

Elwasif, Wael↗
DOE OSTI2024

MassiveGNN: Efficient Training via Prefetching for Massively Connected Distributed Graphs

Graph Neural Networks (GNN) are indispensable in learning from graph-structured data, yet their rising computational costs, especially on massively connected graphs, pose significant challenges in terms of execution performance. To tackle this, distributed-memory solutions such as partitioning the graph to concurrently train multiple replicas of GNNs are in practice. However, approaches requiring a partitioned graph usually suffer from communication overhead and load imbalance, even under optimal partitioning and communication strategies due to irregularities in the neighborhood minibatch sampling. This paper proposes practical trade-offs for improving the sampling and communication overheads for representation learn- ing on distributed graphs (using popular GraphSAGE architecture) by developing a parameterized prefetch and eviction scheme on top of the state-of-the-art Amazon DistDGL distributed GNN framework, demonstrating about 15–40% improvement in end-to-end training performance on the NERSC Perlmutter supercomputer for various OGB datasets.

Machine Leanring, high performance comptuing, grap↗
DOE OSTI2021

Performance Evaluation of Adaptive Routing on Dragonfly-based Production Systems

Performance of applications in production environments can he sensitive to network congestion. Cray Aries supports adaptively routing each network packet independently based on the load or congestion encountered as a packet traverses the network. Software can dictate different routing policies, adjusting between minimal and non-minimal bias, for each posted message. We have extensively evaluated the sensitivity of the routing bias selection on application performance as well as whole system performance in both production and controlled conditions. We show that the default routing bias used in Aries-based systems is often sub-optimal and that using a higher bias towards minimal routes will not only reduce the congestion effects on the application but also will decrease the overall congestion on the network. This routing scheme results in not only improved mean performance (by up to 12%) of most production applications hut also reduced run-to-run variability. Our study prompted the two supercomputing facilities (ALCF and NERSC) to change the default routing mode on their Aries-based systems. We present the substantial improvement measured in the overall congestion management and interconnect performance in production after making this change.

Chunduri, Sudheer↗
DOE OSTI2026

Toward Energy-Efficient HPC: Insights from Power Profiling a Cloud-Resolving Earth System Model

Power is a fundamental constraint as supercomputing advances to exascale. Efficient operation within strict power budgets requires application-aware power management based on a detailed understanding of application-level power behavior. This work analyzes the Energy Exascale Earth System Model (E3SM) atmosphere component, SCREAM, on Perlmutter (NERSC) and Frontier (OLCF). We characterize power variation across inputs, concurrency levels, and power caps, evaluate the energy impact of code optimizations, and attribute energy within the code using a newly developed GPU energy model. Results show that SCREAM’s peak power remains stable during its core execution phase and decreases gradually as concurrency increases. Power capping experiments reveal a performance–energy "sweet spot". On Perlmutter, limiting GPU power to 50% of thermal design power (TDP) achieves up to 15% energy savings with a 7% performance penalty. On Frontier, a 40% TDP cap yields up to 10% energy savings with less than 10% performance loss. Code optimizations reduce SCREAM energy by shortening run time without increasing power. Modeling reveals a critical insight: data movement accounts for approximately 70% of SCREAM’s GPU energy. This fundamentally shifts the optimization focus from FLOPS to data transfer reduction for this class of applications, offering the most impactful strategy for improving energy efficiency. This work establishes a foundation for practical, application-aware power management at exascale.

Zhao, Zhengji [Lawrence Berkeley National Laborato↗
DOE OSTI2021

Interactive Supercomputing With Jupyter

Rich user interfaces like Jupyter have the potential to make interacting with a supercomputer easier and more productive, consequently attracting new kinds of users and helping to expand the application of supercomputing to new science domains. For the scientist-user, the ideal rich user interface delivers a familiar, responsive, introspective, modular, and customizable platform upon which to build, run, capture, document, re-run, and share analysis workflows. From the provider or system administrator perspective, such a platform would also be easy to configure, deploy securely, update, customize, and support. Jupyter checks most if not all of these boxes. But from the perspective of leadership computing organizations that provide supercomputing power to users, such a platform should also make the unique features of a supercomputer center more accessible to users and more composable with high performance computing (HPC) workflows. Project Jupyter’s core design philosophy of extensibility, abstraction, and agnostic deployment, has allowed HPC centers like NERSC to bring in advanced supercomputing capabilities that can extend the interactive notebook environment. This has enabled a rich scientific discovery platform, particularly for experimental facility data analysis and machine learning problems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

GPU Direct I/O with HDF5

Exascale HPC systems are being designed with accelerators, such as GPUs, to accelerate parts of applications. In machine learning workloads as well as large-scale simulations that use GPUs as accelerators, the CPU (or host) memory is currently used as a buffer for data transfers between GPU (or device) memory and the file system. If the CPU does not need to operate on the data, then this is sub-optimal because it wastes host memory by reserving space for duplicated data. Furthermore, this “bounce buffer” approach wastes CPU cycles spent on transferring data. A new technique, NVIDIA GPUDirect Storage (GDS), can eliminate the need to use the host memory as a bounce buffer. Thereby, it becomes possible to transfer data directly between the device memory and the file system. This direct data path shortens latency by omitting the extra copy and enables higher-bandwidth. To take full advantage of GDS in existing applications, it is necessary to provide support with existing I/O libraries, such as HDF5 and MPI-IO, which are heavily used in applications. In this paper, we describe our effort of integrating GDS with HDF5, the top I/O library at NERSC and at DOE leadership computing facilities. We design and implement this integration using a HDF5 Virtual File Driver (VFD). The GDS VFD provides a file system abstraction to the application that allows HDF5 applications to perform I/O without the need to move data between CPUs and GPUs explicitly. We compare performance of the HDF5 GDS VFD with explicit data movement approaches and demonstrate superior performance with the GDS method.

Ravi, J↗
DOE OSTI2024

MDLoader: A Hybrid Model-Driven Data Loader for Distributed Graph Neural Network Training

Scalable data management is essential for processing large scientific dataset on HPC platforms for distributed deep learning. In-memory distributed storage is preferred for its speed, enabling rapid, random, and frequent data access required by stochastic optimizers. Processes use one-sided or collective communication to fetch remote data, with optimal performance depending on (i) dataset characteristics, (ii) training scale, and (iii) interconnection network. Empirical analysis shows collective communication excels with larger mini-batch sizes and/or fewer processes, whereas one-sided communication outperforms at larger scales. We propose MDLoader, a hybrid in-memory data loader for distributed graph neural network training. MDLoader features a model-driven performance estimator that dynamically selects between one-sided and collective communication at the beginning of training using Tree of Parzen Estimators (TPE). Evaluations on NERSC Perlmutter and OLCF Summit show MDLoader outperforms single-backend loaders by up to 2.83 × and predicts the suitable communication method with 96.3% (Perlmutter) and 94.3% (Summit) success rate.

Bae, Jonghyun↗
DOE OSTI2024

Optimizing the Weather Research and Forecasting Model with OpenMP Offload and Codee

Currently, the Weather Research and Forecasting model (WRF) utilizes shared memory (OpenMP) and distributed memory (MPI) parallelisms. To take advantage of GPU resources on the Perlmutter supercomputer at NERSC, we port parts of the computationally expensive routine Fast Spectral Bin Microphysics (FSBM) to NVIDIA GPUs using OpenMP device offloading directives. To facilitate this process, we explore a workflow for optimization which uses both runtime profilers and a static code inspection tool Codee to refactor the subroutine. We observe an 2.24x overall speedup for the CONUS-12km storm test case.

Wichitrnithed, Chayanon (Namo) [Odin Institute]↗
DOE OSTI2024

Enabling Low-Overhead HT-HPC Workflows at Extreme Scale using GNU Parallel

GNU Parallel is a versatile and powerful tool for process parallelization widely used in scientific computing. This paper demonstrates its effective application in high-performance computing (HPC) environments, particularly focusing on its scalability and efficiency in executing large-scale high-throughput high-performance computing (HT-HPC) workflows. Through real-world examples, we highlight GNU Parallel’s performance across various HPC workloads, including GPU computing, container-based workloads, and node-local NVMe storage. Our results on two leading supercomputers, OLCF’s Frontier and NERSC’s Perlmutter, showcase GNU Parallel’s rapid process dispatching ability and its capacity to maintain low overhead even at extreme scales. We explore GNU Parallel’s application in massive parallel file transfers using a scheduled Data Transfer Node (DTN) cluster, emphasizing its broad utility in diverse scientific workflows. Beyond its direct application as a viable workflow manager, GNU Parallel can be employed in conjunction with other workflow systems as a "last-mile" parallelizing driver and as a quick prototyping tool to design and extract parallel profiles from application executions. We then argue that the potential for GNU Parallel to transform workflow management at extreme scales is substantial, paving the way for more efficient and effective scientific discoveries.

Maheshwari, Ketan↗
DOE OSTI2020

Cross-facility science with the Superfacility Project at LBNL

As data sets from DOE user science facilities grow in both size and complexity there is an urgent need for new capabilities to transfer, analyze and manage the data underlying scientific discoveries. LBNL's Superfacility project brings together experimental and observational research instruments with computational and network facilities at the National Energy Research Scientific Computing Center (NERSC) and the Energy Sciences Network (ESnet) with the goal of enabling user science. Here, we report on recent innovations in the Superfacility project, including advanced data management, API-based automation, real-Time interactive user interfaces, and supported infrastructure for 'edge' services.

Enders, B↗
DOE OSTI2021

Machine learning-based analysis of COVID-19 pandemic impact on US research networks

Here in this study we explore how fallout from the changing public health policy around COVID-19 has changed how researchers access and process their science experiments. Using a combination of techniques from statistical analysis and machine learning, we conduct a retrospective analysis of historical network data for a period around the stay-at-home orders that took place in March 2020. Our analysis takes data from the entire ESnet infrastructure to explore DOE high-performance computing (HPC) resources at OLCF, ALCF, and NERSC, as well as User sites such as PNNL and JLAB. We look at detecting and quantifying changes in site activity using a combination of t-Distributed Stochastic Neighbor Embedding (t-SNE) and decision tree analysis. Our findings bring insights into the working patterns and impact on data volume movements, particularly during late-night hours and weekends.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

A Case For Intra-rack Resource Disaggregation in HPC

The expected halt of traditional technology scaling is motivating increased heterogeneity in high-performance computing (HPC) systems with the emergence of numerous specialized accelerators. As heterogeneity increases, so does the risk of underutilizing expensive hardware resources if we preserve today’s rigid node configuration and reservation strategies. This has sparked interest in resource disaggregation to enable finer-grain allocation of hardware resources to applications. However, there is currently no data-driven study of what range of disaggregation is appropriate in HPC. To that end, we perform a detailed analysis of key metrics sampled in NERSC’s Cori, a production HPC system that executes a diverse open-science HPC workload. In addition, we profile a variety of deep-learning applications to represent an emerging workload. We show that for a rack (cabinet) configuration and applications similar to Cori, a central processing unit with intra-rack disaggregation has a 99.5% probability to find all resources it requires inside its rack. In addition, ideal intra-rack resource disaggregation in Cori could reduce memory and NIC resources by 5.36% to 69.01% and still satisfy the worst-case average rack utilization.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Performance Analysis of Speculative Parallel Adaptive Local Timestepping for Conservation Laws

Stable simulation of conservation laws, such as those used to model fluid dynamics and plasma physics applications, requires the satisfaction of the so-called Courant-Friedrichs-Lewy condition. By allowing regions of the mesh to advance with different timesteps that locally satisfy this stability constraint, significant work reduction can be attained when compared to a time integration scheme using a single timestep size. However, parallelizing this algorithm presents considerable difficulty. Since the stability condition depends on the state of the system, dependencies become dynamic and potentially non-local. In this article, we present an adaptive local timestepping algorithm using an optimistic (Timewarp-based) parallel discrete event simulation. We introduce waiting heuristics to limit misspeculation and a semi-static load balancing scheme to eliminate load imbalance as parts of the mesh require finer or coarser timesteps. Last, we outline an interface for separating the physics of the specific conservation law from the temporal integration allowing for productive adoption of our proposed algorithm. We present a misspeculation study for three conservation laws, demonstrating both the productivity of the local timestepping API, for which 74% of the lines of code are reused across different conservation laws, and the robustness of the waiting heuristics—at most 1.5% of element updates are rolled back. Our performance studies demonstrate up to a 2.8× speedup versus a baseline unoptimized local timestepping approach, a 4x improvement in per-node throughput compared to an MPI parallelization of synchronous timestepping, and scalability up to 3,072 cores on NERSC’s Cori Haswell partition.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Latency and Bandwidth Microbenchmarks of US Department of Energy Systems in the June 2023 Top 500 List

As a rule, Top 500 class supercomputers are extensively benchmarked as part of their acceptance testing process. However, barring publicly posted LINPACK / HPCG results, most benchmark results are often inaccessible outside the hosting institution. Moreover, these higher level benchmarks do not provide easy answers to common questions such as “What is the realizable memory bandwidth?” or “What is the launch latency on the accelerator?” To partially address these issues, we executed selected single-node micro-benchmarks — focused on latencies and memory bandwidth — on every US Department of Energy system above rank 150 of the June 2023 Top 500 list, excepting NERSC’s Cori and ORNL’s Frontier TDS (now decommissioned or repurposed). We hope to provide an easy “first stop” reference for users of current Top 500 systems and inspire users and administrators of other Top 500 systems to similarly compile and make available benchmark results for their systems.

Siefert, Christopher M.↗
DOE OSTI2023

Performance-Portable GPU Acceleration of the EFIT Tokamak Plasma Equilibrium Reconstruction Code

This paper presents the steps followed to GPU-offload parts of the core solver of EFIT-AI, an equilibrium reconstruction code suitable for tokamak experiments and burning plasmas. For this work, we will focus on the fitting procedure that consists of a Grad–Shafranov (GS) equation inverse solver that calculates equilibrium reconstructions on a grid. We will show profiling results of the original code (CPU-baseline), as well as the directives used to GPU-offload the most time-consuming function, initially to compare OpenACC and OpenMP on NVIDIA and AMD GPUs and later on to assess OpenMP performance portability on NVIDIA, AMD and Intel GPUs. We will make a performance comparison for different spatial grid sizes and show the speedup achieved on NVIDIA A100 (Perlmutter-NERSC), AMD MI250X (Frontier-OLCF) and Intel PVC GPUs (Sunspot-ALCF). Finally, we will draw some conclusions and recommendations to achieve high-performance portability for an equilibrium reconstruction code on the new HPC architectures

GPU↗
DOE OSTI2021

Customs: Inspect and Report Python Imports (customs) v1.0.0

We have developed a simple, minimally invasive monitoring framework that enables us to capture key metrics from Python user processes. This is part of a larger effort to collect data across our entire data-intensive science workload at NERSC. The approach for Python is inspired by a similar framework developed for Blue Waters at NCSA. Both frameworks leverage standard Python features, sitecustomize and atexit, to capture Python imports of interest and other job data. The customs package decomposes the information capture into separate inspection and reporting interfaces.

Thomas, Rollin↗