Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “memory efficiency”

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 199 records · Page 11
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Parallelization and Performance Portability in Hydrodynamics Codes

With the eve of Exascale computing, performance and portability are at the forefront of all scientific codes. Adding more cores and more energy to a system is no longer a sustainable way to achieve performance, and extra effort must now be made to improve performance in all areas of code and code development. Using hydrodynamic codes as a basis, this work explores numerous techniques to achieve performance in different ways. Adaptive mesh refinement (AMR) is a necessary technique to improve memory optimization in mesh-based simulations. However it is invasive and conventionally difficult to integrate into existing applications, so we present a new branch of AMR to create a smooth transition to these optimizations, which not only improves performance, but also greatly reduces developer effort. We introduce the concept of this improvement as Phantom-Cell AMR, and assess theoretically the improvements, as well as present an application of its use. Other work included involves and investigation into an efficient data structure that ensures optimal memory layout for cache performance, with a target of making codes performant and portable across all architectures. All of the work targets both performance and portability, not just on CPU hardware, but specifically across GPU architectures. Parallel performance is key to all of the methods presented, but the research makes a great effort to improve the portability of all applications to prepare for current high performance computing systems and those on the horizon.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Compute in‐Memory with Non‐Volatile Elements for Neural Networks: A Review from a Co‐Design Perspective

Abstract Deep learning has become ubiquitous, touching daily lives across the globe. Today, traditional computer architectures are stressed to their limits in efficiently executing the growing complexity of data and models. Compute‐in‐memory (CIM) can potentially play an important role in developing efficient hardware solutions that reduce data movement from compute‐unit to memory, known as the von Neumann bottleneck. At its heart is a cross‐bar architecture with nodal non‐volatile‐memory elements that performs an analog multiply‐and‐accumulate operation, enabling the matrix‐vector‐multiplications repeatedly used in all neural network workloads. The memory materials can significantly influence final system‐level characteristics and chip performance, including speed, power, and classification accuracy. With an over‐arching co‐design viewpoint, this review assesses the use of cross‐bar based CIM for neural networks, connecting the material properties and the associated design constraints and demands to application, architecture, and performance. Both digital and analog memory are considered, assessing the status for training and inference, and providing metrics for the collective set of properties non‐volatile memory materials will need to demonstrate for a successful CIM technology.

36 MATERIALS SCIENCE↗
DOE OSTI2025

NLR HPC Eagle Jobs Data and Additional Energy Metrics

Overview: Anonymized job-level records from the Eagle high-performance computing (HPC) system at the National Laboratory of the Rockies (NLR). Each record represents a Slurm batch job with scheduling metadata, resource requests, resource utilization, CPU/GPU energy consumption, and efficiency metrics. Sensitive fields (user, account, job name) are replaced with cryptographic hashes. System & Timeframe: Eagle was a 2,000-node, 8-petaflop system operated at NLR from 2019–2024. Data covers the full operational lifetime of the system. Slurm data was processed nightly; timestamps are in Mountain Time. Funding provided by the U.S. Department of Energy, EERE. Files: esif.hpc.eagle.job-anon.zip — Core anonymized job records (Hive-partitioned Parquet) esif.hpc.eagle.job-anon-energy-metrics.zip — Same records with additional iLO and Ganglia energy metrics datacard.md — Full dataset documentation ~13.8 million rows, 62 variables. Readable with PyArrow, pandas, DuckDB, Apache Spark, or any Parquet-compatible tool. Data Collection: Jobs collected via sacct through a pipeline: Eagle Jobs API → Redpanda → StreamSets → HPCMON API → PostgreSQL. Node-level power from iLO (HP Integrated Lights-Out); GPU power from Ganglia monitoring, joined to jobs via node lists and time ranges. Preprocessing: Anonymization of name, user, and account fields via cryptographic hashing Derived columns: queue_wait, cpu_eff, max_mem_eff Simplified job state mapping (e.g., "CANCELLED BY 12345" → "CANCELLED") QoS accounting rules (buy-in, standby, or Slurm QoS value) CPU energy estimated from TDP (200W, Intel Xeon Gold 6154, 18 cores) Timezone-aware columns (_tz) sourced from LEX accounting database to correctly handle DST transitions Key Variables: Scheduling: job_id, partition, state_simple, submit_time_tz, start_time_tz, end_time_tz, queue_waitResources: nodes_req/used, processors_req/used, memory_req, wallclock_req/used, gpus_requested Efficiency: cpu_eff, max_mem_eff Energy: cpu_energy_tdp_estimated_max/used_watt_hours, node_energy_total_watt_hours (iLO), gpu0/1_energy_total_watt_hours (Ganglia) Partitions: bigmem, bigmem-8600, bigscratch, csc, dav, ddn, debug, gpu, haswell, long, mono, short, standard Job States: CANCELLED, COMPLETED, FAILED, NODE_FAIL, OUT_OF_MEMORY, PENDING, RUNNING, TIMEOUT QoS Levels: Unknown, normal, buy-in, debug, penalty, high, standby Important Notes: Non-_tz timestamp columns may be off by one hour across DST boundaries; use _tz columns for time difference calculations Energy fields are null for jobs without monitoring coverage Job step records and raw Slurm JSONB fields are excluded from this extract Do not attempt to re-identify individuals from hashed fields

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

XploreNAS : Explore Adversarially Robust and Hardware-efficient Neural Architectures for Non-ideal Xbars

Compute In-Memory platforms such as memristive crossbars are gaining focus as they facilitate acceleration of Deep Neural Networks (DNNs) with high area and compute efficiencies. However, the intrinsic non-idealities associated with the analog nature of computing in crossbars limits the performance of the deployed DNNs. Furthermore, DNNs are shown to be vulnerable to adversarial attacks leading to severe security threats in their large-scale deployment. Thus, finding adversarially robust DNN architectures for non-ideal crossbars is critical to the safe and secure deployment of DNNs on the edge. This work proposes a two-phase algorithm-hardware co-optimization approach called XploreNAS that searches for hardware efficient and adversarially robust neural architectures for non-ideal crossbar platforms. We use the one-shot Neural Architecture Search approach to train a large Supernet with crossbar-awareness and sample adversarially robust Subnets therefrom, maintaining competitive hardware efficiency. Our experiments on crossbars with benchmark datasets (SVHN, CIFAR10, CIFAR100) show up to ~8–16% improvement in the adversarial robustness of the searched Subnets against a baseline ResNet-18 model subjected to crossbar-aware adversarial training. We benchmark our robust Subnets for Energy-Delay-Area-Products (EDAPs) using the Neurosim tool and find that with additional hardware efficiency–driven optimizations, the Subnets attain ~1.5–1.6× lower EDAPs than ResNet-18 baseline.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Characterization of throughput on the AXI DMA bus for burst data transfer over Ethernet

cThe Xilinx AXI Direct Memory Access (AXI DMA) module is an efficient solution for medium-speed data transfer in Xilinx SoC FPGAs, supporting data rates greater than 1000 Gbps even in very suboptimal operating modes. It facilitates direct transfer of AXI stream data into processor memory without constant software intervention, which reduces overhead and ensures consistent data logging. By utilizing the FPGA's available memory, large circular buffers (1-5 GiB) are used to buffer data and accommodate network limitations, enabling high-rate data bursts. In this study, we measured the performance of AXI DMA under conditions simulating its lowest practical data transfer speeds. The Arbitrary Length Data Sender was used to transmit AXI stream packets at 32-bit width and 100 MHz frequency, a narrow width and slow speed. Results show that the AXI DMA can transfer up to 3192.76 Mbps with large packet sizes but experiences reduced performance for smaller packets, as low as 2.6 Mbps for 4-byte packets. For Ethernet-limited applications, packet sizes between 8,000 and 16,000 bytes provided optimal transfer speeds of 874 to 1600 Mbps. These findings suggest that the AXI DMA is not the limiting factor in systems where packet sizes exceed 8,000 bytes.

43 PARTICLE ACCELERATORS↗
DOE OSTI2025

NLR HPC Kestrel Jobs Data

Overview: Anonymized job-level records from the Kestrel HPC system at the National Laboratory of the Rockies (NLR). Each record represents a Slurm batch job with scheduling metadata, resource requests, utilization, energy estimates, and efficiency metrics. Sensitive fields (user, account, job name, submit line, working directory, submit script, and job type) are replaced with 7-character cryptographic hashes. System & Timeframe: Kestrel is located at the NLR campus. Standard compute nodes have 104 cores and 256 GB RAM; bigmem nodes have 2,000 GB. GPU nodes (gpu-h100 partition) use NVIDIA H100 GPUs. Data covers jobs submitted August 2023 through December 2025. Funding provided by the U.S. Department of Energy, EERE. Files: esif.hpc.kestrel.job-anon.zip — Anonymized job records (Hive-partitioned Parquet) datacard.md — Full dataset documentation ~11 million rows, 50 variables. Readable with PyArrow, pandas, DuckDB, Apache Spark, or any Parquet-compatible tool. Data Collection: Jobs collected via sacct with timezone-aware export (SLURM_TIME_FORMAT="%Y-%m-%dT%H:%M:%S%z"), loaded into PostgreSQL. Calculated columns updated via database triggers and batch functions. All timestamps use timestamptz and correctly handle DST transitions. Preprocessing: Anonymization of name, user, account, submit_line, work_dir, submit_script, and job_type via 7-char hex hashes Derived columns: queue_wait, cpu_eff, max/min/avg_mem_eff, energy estimates Simplified job state mapping (e.g., "CANCELLED by 132357" → "CANCELLED") Boolean flags: python_job, reframe_job Temporal decomposition: year, month, day, day_of_week, hour, minute from submit_time Shared node tracking: shared_job_count, nodes_shared, jobs_shared Key Variables: Scheduling: job_id, partition, state_simple, submit_time, start_time, end_time, queue_wait Resources: nodes_req/used, processors_req/used, memory_req, wallclock_req/used, gpus_requested Efficiency: cpu_eff, max/min/avg_mem_eff Energy: cpu_energy_tdp_estimated_max/used_watt_hours, consumed_energy_raw_joules, consumed_energy_raw_watt_hours Sharing: shared_job_count, nodes_shared, jobs_shared Partitions: short, standard, debug, gpu-h100 Job States: CANCELLED, COMPLETED, FAILED, PENDING, RUNNING QoS Levels: normal, high Important Notes: Timestamps include timezone offsets; DST transitions are handled correctly, though adding intervals across DST boundaries requires offset adjustment shared_job_count reflects physical node co-residency, not use of the shared partition Job step records and raw Slurm JSONB fields are excluded Do not attempt to re-identify individuals from hashed fields

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Effect of Gamma Radiation on TaOₓ ECRAM

Electrochemical random access memory (ECRAM) is an emerging three-terminal nonvolatile memory (NVM) with highly controllable channel conductance which is promising for use as an analog memory (or synapse) in analog in-memory computing (IMC) systems. Energy-efficient analog IMC computing is particularly desirable for power-constrained, high-radiation environments such as satellites. However, little is known about the suitability of ECRAM for use in a total ionizing dose (TID) environment. Here, this work investigates the effect of Co-60 gamma radiation on the channel conductance and noise—two properties critical for analog IMC systems—of a TaO x -based ECRAM up to 17.3 Mrad(SiO 2 ) for both low- and high-channel-conductance state devices. A transient increase in conductance is observed in response to radiation which consists of two elements: an immediate increase in conductivity due to photocurrent and a secondary increase in conductivity, which has a slower rise and saturation and can persist for hours after exposure. This secondary, persistent photoconductivity is attributed to charging caused by hole trapping. These transient effects would not likely occur in a space environment due to the low dose rate compared with this experiment. No permanent change is found in the low conductance state (LCS) following exposure and the minor shift in the high conductance change would be less significant than the regular retention decay in this state. A permanent increase in the random telegraph noise is observed, possibly due to increased traps created in the channel. This work demonstrates that TaO x -based ECRAM is suitable for use in spaceborne analog IMC systems that are subject to significant TID.

ECRAM↗
DOE OSTI2025

Purcell-Enhanced Emissions from Diamond Color Centers in Slow Light Photonic Crystal Waveguides

Diamond color centers are promising candidates for optically addressable quantum memories, which motivates the development of efficient photonic interfaces, often using nanophotonic cavities with narrow spectral line widths and small mode volumes. However, they require perfect spectral and spatial overlap between the cavity mode and quantum emitter, which is challenging. This is especially true for solid-state quantum emitters that are often randomly positioned and suffer from inhomogeneous broadening. Another approach to enhance light-matter interaction across large optical bandwidths and areas is using slow light waveguides. Here, in this study, we demonstrate diamond slow light photonic crystal (PhC) waveguides optically coupled to embedded silicon-vacancy (SiV) color centers. We use the recently developed thin-film diamond approach to fabricate fully suspended two-dimensional PhC waveguides. We demonstrate waveguide modes with high group indices up to 70 and observe Purcell-enhanced emissions of the SiVs. Our approach represents a practical diamond platform for robust spin-photon interfaces with color centers.

2D photonic crystal waveguide↗
DOE OSTI2025

S AP F LOWER : an automated tool for sap flow data preprocessing, gap-filling, and analysis using deep learning

Sap flow, a critical process in plant water use and ecosystem water cycles, is often measured using thermal dissipation probes (TDP) due to their ease of installation and continuous data collection. However, sap flow data frequently include noise, outliers, and gaps, creating challenges for analysis and requiring substantial manual processing. We developed S AP F LOWER , a tool that automates data preprocessing, model training, gap-filling, sapwood area scaling and modeling, and water use analysis. It integrates autocleaning, machine learning and deep learning models (e.g. random forest, Gaussian process regression, long short-term memory (LSTM), bidirectional LSTM (BiLSTM)), and efficient workflows to process sap flow data. S AP F LOWER can remove over 90% of noisy data while preserving legitimate variations and achieve high accuracy in gap-filling based on user-determined parameters. Random forest, LSTM, and BiLSTM models reduced root mean square error to 10% or less for long-term gaps. Model training and prediction can be performed efficiently within seconds. S AP F LOWER significantly enhances the efficiency and accessibility of TDP data analysis by automating complex tasks, enabling researchers without programming expertise to employ advanced techniques. Future improvements will focus on species-specific corrections for TDP and support for additional measurement methods. S AP F LOWER is openly available on GitHub (https://github.com/JiaxinWang123/SapFlower) and Zenodo (doi: 10.5281/zenodo.13665919).

ecosystem water balance↗
DOE OSTI2021

PLANC: Parallel Low-rank Approximation with Nonnegativity Constraints

In this work, we consider the problem of low-rank approximation of massive dense nonnegative tensor data, for example, to discover latent patterns in video and imaging applications. As the size of data sets grows, single workstations are hitting bottlenecks in both computation time and available memory. We propose a distributed-memory parallel computing solution to handle massive data sets, loading the input data across the memories of multiple nodes, and performing efficient and scalable parallel algorithms to compute the low-rank approximation. We present a software package called Parallel Low-rank Approximation with Nonnegativity Constraints, which implements our solution and allows for extension in terms of data (dense or sparse, matrices or tensors of any order), algorithm (e.g., from multiplicative updating techniques to alternating direction method of multipliers), and architecture (we exploit GPUs to accelerate the computation in this work). We describe our parallel distributions and algorithms, which are careful to avoid unnecessary communication and computation, show how to extend the software to include new algorithms and/or constraints, and report efficiency and scalability results for both synthetic and real-world data sets.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

DIMPLES: Distributed Influence Maximization for Pandemic pLanning on Exascale Systems

We study exascale parallel algorithms for the selection of intervention or monitoring strategies in massive realistic socio-technical networks through scalable Influence Maximization (InfMax) algorithms. We employ novel techniques to enable efficient scaling on up to 8k nodes of OLCF Frontier, with 65k AMD GPUs and 458k AMD CPU cores. Current state-of-the-art InfMax tools are limited to networks with only a few million actors (vertices) and a few hundred million interactions (edges). By overcoming these limitations, we show that our approach is capable of processing a realistic social contact network of the United States with 285 million nodes and about 8 billion edges. This two orders-of-magnitude improvement over the previous state-of-the-art is obtained by leveraging algorithmic advancements for the InfMax problem and designing several problem-specific approaches to overlap communication with computation, improve GPU efficiency, and lower the application’s memory requirements. We evaluate strong scaling for computing 10k most influential seeds using up to 8k nodes of an exascale system, and weak scaling from 128 to 8k system nodes for seed sets ranging from 625 to 40k seeds. We achieve the fastest-known runtime of 25 minutes while performing 48 million diffusion simulations totaling 2.31 petabytes to identify 40k influential seeds using 8k nodes, and take 5.75 minutes to identify 10k seeds while using 4k nodes.

Minutoli, Marco [Pacific Northwest National Labora↗
DOE OSTI2021

pyDRESCALk

Modern data scientists are tasked to analyze ever-growing data sets with increasingly complex relationships. Tensor decompositions have come to play a central role in identifying underlying latent structures in higher-order data. The problem of fitting tensor models to different distributions is complicated by the combinations of size, dimensionality, and sparsity present in real world data. The situation demands efficient algorithms designed for shared-memory and distributed systems. This work will present new research that tackles these challenges on several different fronts, leveraging optimizations in numerical algorithms and sparse tensor representations in heterogeneous high performance computing environments.

Bhattarai, Manish↗
DOE OSTI2026

Chemically Enabled CO 2 -Enhanced Oil Recovery in Multi-Porosity, Hydrothermally Altered Carbonates in the Southern Michigan Basin (Final Technical Report)

This is the Final Technical Report for the project "Chemically Enabled CO 2 -Enhanced Oil Recovery in Multi-Porosity, Hydrothermally Altered Carbonates in the Southern Michigan Basin." Over the course of six years of collaboration between Battelle and project partners, all stated objectives of the program have been completed, including full geological characterization of the TBR trend (See companion report for Task2), laboratory and modeling experiments to determine the optimum composition and design of CO 2 -EOR operations in the TBR trend, execution of a field test of chemically-enhanced CO 2 in a TBR well, and integration of the data and learnings gathered during these efforts into a full-trend development plan. Detailed reporting on these activities, their outcomes, and implications for trend-wide development is provided in the report. This report and encompassed data will provide TBR field operators with detailed information on what worked, what did not work, and how to proceed with production optimization of their TBR assets using chemically-enhanced CO 2 -EOR. CO 2 -EOR is a relatively well understood and broadly implemented strategy for increasing incremental production across the oil and gas industry, but its application has been primarily focused on reservoirs with limited heterogeneity. The intention of this project was show first that the same physical mechanisms that improve recovery factors in homogeneous reservoirs (namely wettability alteration, viscosity alteration, oil swelling, and mobility control) are at play in heterogeneous reservoirs. This was proven by the project’s laboratory studies and dynamic simulations, with the potential exception of mobility control, which needs further study. The second intention was to demonstrate via direct field testing that CO 2 -EOR can work in a strongly heterogeneous reservoir. While the field test strategy implemented during this project did not succeed in producing oil, data gathered during the test sheds light on what may work for field operators who try chemically-enhanced CO 2 -EOR within their own reservoirs, significantly reducing the level of uncertainty carried by first-of-a-kind commercial efforts that could (and should) follow this test. Simultaneously, the project has identified several large-volume ethanol plants and other sources of CO 2 emissions in the region and provided a handrail that CO 2 emitters and field operators can leverage to capture, transport, and inject that CO 2 into their fields. This project has also shown that, in many cases, the economics of CO 2 -EOR in the TBR are attractive. And finally, by completing a project of this scope in the southern Michigan Basin, the project has contributed to the knowledge base and operational experience of field operators, state regulatory agencies, local service companies, and state universities, with CO 2 -EOR projects which should allow follow-on projects to proceed safely and efficiently.

02 PETROLEUM↗
DOE OSTI2024

Enhancing Lattice Kinetic Schemes for Fluid Dynamics with Lattice-Equivariant Neural Networks

A new class of equivariant neural networks is presented, hereby dubbed lattice-equivariant neural networks (LENNs), designed to satisfy local symmetries of a lattice structure. The approach develops within a recently introduced framework aimed at learning neural network-based surrogate models’ lattice Boltzmann collision operators. Whenever neural networks are employed to model physical systems, respecting symmetries and equivariance properties has been shown to be key for accuracy, numerical stability, and performance. Here, hinging on ideas from group representation theory, trainable layers are defined whose algebraic structure is equivariant with respect to the symmetries of the lattice cell. In this work, the presented method naturally allows for efficient implementations, in terms of both memory usage and computational costs, supporting scalable training/testing for lattices in two spatial dimensions and higher (in which the size of symmetry group grows). The approach is validated and tested considering 2D and 3D flowing dynamics, both in laminar and turbulent regimes. It is compared with group-averaged-based symmetric networks and with plain, nonsymmetric, networks, showing how the presented approach unlocks the (a posteriori) accuracy and training stability of the former models and the train/inference speed of the latter networks. (LENNs are about one order of magnitude faster than group-averaged networks in 3D.) The work in this paper opens toward practical use of machine learning-augmented lattice Boltzmann CFD in real-world simulations.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

UPC++ as_eager Working Group Draft, Revision 2020.6.2

This draft proposes an extension for a new future-based completion variant that can be more effectively streamlined for RMA and atomic access operations that happen to be satisfied at runtime using purely node-local resources. Many such operations are most efficiently performed synchronously using load/store instructions on shared-memory mappings, where the actual access may only require a few CPU instructions. In such cases we believe it’s critical to minimize the overheads imposed by the UPC++ runtime and completion queues, in order to enable efficient operation on hierarchical node hardware using shared-memory bypass. The new upcxx::{source,operation}_cx::as_eager_future() completion variant accomplishes this goal by relaxing the current restriction that future-returning access operations must return a non-ready future whose completion is deferred until a subsequent explicit invocation of user-level progress. This relaxation allows access operations that are completed synchronously to instead return a ready future, thereby avoiding most or all of the runtime costs associated with deferment of future completion and subsequent mandatory entry into the progress engine. We additionally propose to make this new as_eager_future() completion variant the new default completion for communication operations that currently default to returning a future. This should encourage use of the streamlined variant, and may provide performance improvements to some codes without source changes. A mechanism is proposed to restore the legacy behavior on-demand for codes that might happen to rely on deferred completion for correctness. Finally, we propose a new as_eager_promise() completion variant that extends analogous improvements to promise-based completion, and corresponding changes to the default behavior of as_promise().

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

A massively parallel and scalable multi-CPU material point method

Harnessing the power of modern multi-GPU architectures, we present a massively parallel simulation system based on the Material Point Method (MPM) for simulating physical behaviors of materials undergoing complex topological changes, self-collision, and large deformations. Our system makes three critical contributions. First, we introduce a new particle data structure that promotes coalesced memory access patterns on the GPU and eliminates the need for complex atomic operations on the memory hierarchy when writing particle data to the grid. Second, we propose a kernel fusion approach using a new Grid-to-Particles-to-Grid (G2P2G) scheme, which efficiently reduces GPU kernel launches, improves latency, and significantly reduces the amount of global memory needed to store particle data. Finally, we introduce optimized algorithmic designs that allow for efficient sparse grids in a shared memory context, enabling us to best utilize modern multi-GPU computational platforms for hybrid Lagrangian-Eulerian computational patterns. We demonstrate the effectiveness of our method with extensive benchmarks, evaluations, and dynamic simulations with elastoplasticity, granular media, and fluid dynamics. In comparisons against an open-source and heavily optimized CPU-based MPM codebase [Fang et al. 2019] on an elastic sphere colliding scene with particle counts ranging from 5 to 40 million, our GPU MPM achieves over 100x per-time-step speedup on a workstation with an Intel 8086K CPU and a single Quadro P6000 GPU, exposing exciting possibilities for future MPM simulations in computer graphics and computational science. Moreover, compared to the state-of-the-art GPU MPM method [Hu et al. 2019a], we not only achieve 2x acceleration on a single GPU but our kernel fusion strategy and Array-of-Structs-of-Array (AoSoA) data structure design also generalizes to multi-GPU systems. Our multi-GPU MPM exhibits near-perfect weak and strong scaling with 4 GPUs, enabling performant and large-scale simulations on a 10243 grid with close to 100 million particles with less than 4 minutes per frame on a single 4-GPU workstation and 134 million particles with less than 1 minute per frame on an 8-GPU workstation.

Wang, Xinlei↗