Search NASA⌕ Search
Home/Archive/DOE OSTI

DOE OSTI · 1798430

Enhancements supporting IC usage of PEM libraries on next-gen platforms

Richards, David F.·Ryujin, Brian S.·Barton, Nathan·Bastea, Sorin·Beauchamp, Brock·Beck, Bret·Beckingsale, David·Blake, Robert·Brantley, Patrick·Brown, Peter·Burmark, Jason·Carson, Robert·Chen, Evelyn·Collette, Mike·Dawson, Shawn·Fried, Larry·Gert, Godfree·Grondalski, John·Gyllenhaal, John·Hall, Burl·Haque, Riyaz·Isaac, Ben·Katz, M·Kunen, Adam·Kuo, I-F Will·Le, Hai·Loffeld, John·Mattoon, Caleb·McFadden, Marty·McKinley, Scott·Meraz-Rodriguez, Manny·Miller, Doug·Minner, Paul·Nimmakayala, Rao·Noble, Chad·O'Brien, Matt·Osawe, Maxwell·Patel, Mehul·Pozulp, Mike·Pudliner, Brian·Rana, Verinder·Rieben, Rob·Robinson, Peter·Skinner, Aaron·Slone, Dale·Stephens, Branson·Sterne, Philip·Stevens, David·Stitt, Tom·Vargas, Arturo·Wayne, Brett·Weiss, Kenneth·White, Chris·Whitesides, Russell·Yang, Max·Yee, Ben
View original source ↗Read source document ↗Export metadata ↓

Abstract

This milestone reports on the culmination of several years of effort by multiple PEM support software development teams to provide capabilities for use in LLNL-developed integrated codes on next-gen ASC platforms, including GPU support. We will provide a survey of relevant Application Program Interfaces (API) that are required to support LLNL IC code capability on relevant architectures, with a focus on Sierra and El Capitan. We will identify and summarize all dependencies between PEM supported libraries and IC supported physics codes. We will provide an assessment of algorithmic improvements that have been deployed, as well as future developments that are required to complete the GPU porting efforts. This assessment will include a description of programming models adopted by each of the PEM projects, distinct algorithmic challenges for each of the capabilities, and information about sharing GPU memory between the APIs and host codes. We will develop targeted test problems to assess computational performance. Finally, this milestone will result in identification of gaps in our effort to assist the LLNL ASC program in prioritization of effort for porting software to El Capitan.

Explore related subjects

97 MATHEMATICS AND COMPUTING

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Richards, David F., Ryujin, Brian S., Barton, Nathan, Bastea, Sorin, Beauchamp, Brock, Beck, Bret, Beckingsale, David, Blake, Robert, Brantley, Patrick, Brown, Peter, Burmark, Jason, Carson, Robert, Chen, Evelyn, Collette, Mike, Dawson, Shawn, Fried, Larry, Gert, Godfree, Grondalski, John, Gyllenhaal, John, Hall, Burl, Haque, Riyaz, Isaac, Ben, Katz, M, Kunen, Adam, Kuo, I-F Will, Le, Hai, Loffeld, John, Mattoon, Caleb, McFadden, Marty, McKinley, Scott, Meraz-Rodriguez, Manny, Miller, Doug, Minner, Paul, Nimmakayala, Rao, Noble, Chad, O'Brien, Matt, Osawe, Maxwell, Patel, Mehul, Pozulp, Mike, Pudliner, Brian, Rana, Verinder, Rieben, Rob, Robinson, Peter, Skinner, Aaron, Slone, Dale, Stephens, Branson, Sterne, Philip, Stevens, David, Stitt, Tom, Vargas, Arturo, Wayne, Brett, Weiss, Kenneth, White, Chris, Whitesides, Russell, Yang, Max, Yee, Ben. 2021-06-23. Enhancements supporting IC usage of PEM libraries on next-gen platforms. https://doi.org/10.2172/1798430

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related reports

DOE OSTI2026

TANTE: Time-adaptive operator learning via neural Taylor expansion

Operator learning for time-dependent partial differential equations (PDEs) has seen rapid progress in recent years, enabling efficient approximation of complex spatiotemporal dynamics. However, most existing methods rely on fixed time step sizes during rollout, which limits their ability to adapt to varying temporal complexity and often leads to error accumulation. In this work, we propose the Time-Adaptive Transformer with Neural Taylor Expansion (TANTE), a novel operator-learning framework that produces continuous-time predictions with adaptive step sizes. TANTE predicts future states by performing a Taylor expansion at the current state, where neural networks learn both the higher-order temporal derivatives and the local radius of convergence. This allows the model to dynamically adjust its rollout based on the local behavior of the solution, thereby reducing cumulative error and improving computational efficiency. We demonstrate the effectiveness of TANTE across a wide range of PDE benchmarks, achieving superior accuracy and adaptability compared to fixed-step baselines, delivering accuracy gains of 60-80 % and speed-ups of 30-40 % at inference time.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Structured illumination for surface-resolved grazing-incidence X-ray scattering

Grazing-incidence (GI) scattering techniques are widely used to characterize thin films, offering high surface sensitivity and insight into morphology and structure. However, these approaches typically provide statistical averaged information due to elongated footprint or limited spatial resolution due to beam size. Here we introduce a method that combines structured illumination with GI X-ray scattering and leverages our computational imaging approach to resolve local structural details. We demonstrate that our method captures local features of an organic semiconductor thin film without the need for sample rotation as in tomography. The method expands GI techniques from statistical averaging to high-resolution imaging, thereby providing the capability for detailed analysis of local material properties, such as domain shape, orientation and polymorphism, which are critical for advancing material design towards more efficient and tailored materials.

97 MATHEMATICS AND COMPUTING↗