High Performance Computing and Communications: Investment in American Competitiveness

In the field of semiconductor quantum dot spin qubits, there is growing interest in leveraging the unique properties of hole-carrier systems and their intrinsically strong spin-orbit coupling to engineer novel spin qubits. In contrast to the weak spin-orbit interaction found in Si, strong spin-orbit coupling in Ge may allow purely electrical spin control using the electric field component of a microwave drive signal applied to a confinement gate, without the need for additional components such as microwave stripline antennas or micromagnets. A crucial aspect in the search for alternative semiconductor qubit host materials is the need to maintain the advantages of existing platforms such as Si: low disorder, long coherence times, extensibility, and semiconductor foundry compatibility. Recent advances in semiconductor heterostructure growth have made available high quality, undoped Ge/SiGe quantum wells that can serve as hosts for hole spin qubits. These quantum wells consist of a pure strained Ge layer flanked by Ge-rich SiGe layers above and below, grown by reduced pressure chemical vapor deposition. Isotopic enrichment is possible if required. The heavy hole charge carriers are characterized by a small effective mass m* ~ 0.08m 0 (about 2.4x lighter than for electrons in Si), high mobility μ ≥ 1E5 cm2/Vs, and highly anisotropic g-factors. The small effective mass should allow relatively large and easily-coupled quantum dots. The next-nearest light hole band is theoretically predicted to be on the order of ~ 100 meV below the heavy hole band. The p-orbital character of the hole wavefunction may provide an advantage in minimizing contact hyperfine interactions with nonzero-spin lattice nuclei. There is both theoretical and experimental evidence of a cubic Rashba-type spin-orbit interaction in this material, whose large strength is manifest as a spin-orbit length on the order of 1 micrometer. An additional advantage over Si is that the two-dimensional band structure of Ge quantum wells does not feature valley states, which can present difficulties for qubit operations. Here, we will describe our ongoing efforts to prepare spin qubits in this platform, including development of gated device architectures, device tuning protocols, and charge-sensing capabilities. Initial work focused on single-layer gate layouts, in which lithographically-defined single and double quantum dots were demonstrated at dilution refrigerator temperatures. Charging energies of ~ 1.7 meV were observed. These devices featured a broad and shallow potential landscape, and only a small number of holes could be confined to a quantum dot. Although adjacent quantum dots could be coupled to one another, this coupling was not easily tuned using this gate design. These limitations in the confinement and control of quantum dots motivated the subsequent adoption of multilayer gate-stack architectures, similar to what has been demonstrated in Si. Our three-layer devices feature gate electrodes designed for accumulation of holes, local depletion to form quantum dots, and selective isolation/screening of electric fields. Device structures feature Ti/Pt gates, ALD oxide dielectrics, and implanted ohmic contacts. Symmetrically constructed devices comprise two gated channels for quantum dot confinement, one to be used as the quantum dot under study and the other operated as a local charge sensor. By monitoring the conductance of the charge sensor as the quantum dot is tuned, we can remotely determine the dot’s occupation number, in principle down to the last hole. Iterative cycles of device fabrication and measurement, combined with numerical modeling, have allowed continuous improvements to the device design. Early three-layer device designs suffered from poor isolation between the upper and lower mirror-symmetric device channels. Numerical modeling showed (and measurements have confirmed) that an additional isolation gate could allow improved autonomy of the two channels while maintaining their electrostatic coupling needed for charge sensing. Ongoing, simulation-informed work to fine-tune the device geometry, as well as efforts toward a qubit demonstration, will be discussed. This work was funded, in part, by the Laboratory Directed Research and Development Program and performed, in part, at the Center for Integrated Nanotechnologies, an Office of Science User Facility operated for the US Department of Energy (DOE) Office of Science. Sandia National Laboratories is a multi-mission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the US Department of Energy’s National Nuclear Security Administration under contract DE-NA-0003525. The views expressed in this article do not necessarily represent the views of the US Department of Energy or the United States Government. The work at NTU was supported by the Ministry of Science and Technology (107-2622-8-002-018-). Figure 1

The U.S. Bioeconomy: Charting a Course for a Resilient and Competitive Future

The HHL algorithm is a prominent quantum algorithm that offers exponential speedup over its classical counterparts for solving a system of linear equations. However, synthesizing and executing HHL circuits demand significant computational resources from both classical and quantum systems. In this paper, we benchmark the HHL algorithm using the optimized Quantum Phase Estimation (QPE) generation algorithm, LuGo [1], to enhance its scalability and efficiency. We leverage various highperformance computing (HPC) systems–the National Energy Research Scientific Computing Center's (NERSC's) Perlmutter supercomputer and Oak Ridge Leadership Computing Facility's (OLCF's) Andes system and Frontier supercomputer–to evaluate the scalability of generating HHL circuits and to measure the time to simulate the generated circuits. Additionally, enabled by the Quantum Computing User Program (QCUP) of OLCF, we provide a comprehensive analysis of the algorithm's performance on various state-of-the-art superconducting and trappedion quantum devices, including studies on qubit connectivity, fidelity comparisons, and hardware compatibility and robustness. Our results offer preliminary insights into potential practical applications of the HHL algorithm enabled by LuGo and the performance of various types of quantum hardware. <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup><sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup>Notice: This manuscript has been authored by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the US Department of Energy (DOE). The US government retains and the publisher, by accepting the article for publication, acknowledges that the US government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US government purposes. DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan). This research used resources of the Oak Ridge Leadership Computing Facility at the Oak Ridge National Laboratory, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC05-00OR22725. This research used resources of the National Energy Research Scientific Computing Center (NERSC), a U.S. Department of Energy Office of Science User Facility located at Lawrence Berkeley National Laboratory, operated under Contract No. DE-AC02-05CH11231.

As announced in the hardware Chap. 5, there is a relatively new paradigm of computation in which a single very large application problem is partitioned into subproblems which can be computed concurrently by being distributed over a network of computers, called a cluster, usually a large-scale local area network (LAN) in which nodes of the network can be connected directly through large switches to build a message-passing system. The given application problem is computationally intensive and would take too long to solve on an existing single computer, however fast it may run. For certain problem structures it is possible to partition the problem into components which can be run concurrently to achieve a speed-up that makes the solution feasible. An example is the solution of a linear system Ax = b where A is a nxn matrix and b is an n-dimensional vector. The Jacobi iterative method can be programmed to work on K groups of m components of x concurrently where n = Km. The K groups are computed on K nodes of the cluster. The individual K computers installed as the nodes in the network, usually pc’s, operate independently on their assigned subproblems for the most part, but at various points in their computations they must communicate with each other by a message-passing protocol which permits the cooperative exchange of intermediate results needed in the various subproblems. This paradigm is designated as high performance computing and communication (HPCC).KeywordsCloud ComputingFile SystemMessage Passing InterfaceHigh Performance ComputingChief Information OfficerThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

This paper describes the emerging concept of High Performance Computing and Communications (HPCC) and discusses its technological, scientific and economic dimensions. It then briefly reviews development trends in the United States and Japan with a view to map the state of the art in Europe. Emphasis is placed on public, national and regional promotion programmes to develop and/or implement HPCC systems for scientific applications.

The operation of large US Department of Energy (DOE) research facilities, like the DIII-D National Fusion Facility, results in the collection of complex multi-dimensional scientific datasets, both experimental and model-generated. In the future, it is envisioned that integrated data analysis coupled with large-scale high performance computing (HPC) simulations will be used to improve experimental planning and operation. Practically, massive data sets from these simulations provide the physics basis for generation of both reduced semi-analytic and machine-learning-based models. Storage of both HPC simulation datasets (generated from US DOE leadership computing facilities) and experimental datasets presents significant challenges. In this paper, we present a vision for a DOE-wide data management workflow that integrates US DOE fusion facilities with leadership computing facilities. Data persistence and long-term availability beyond the length of allocated projects is essential, particularly for verification and recalibration of artificial intelligence and machine learning (AI/ML) models. Because these data sets are often generated and shared among hundreds of users across multiple leadership computing facility centers, they would benefit from cross-platform accessibility, persistent identifiers (e.g. DOI, or digital object identifier), and provenance tracking. The ability to handle different data access patterns suggests that a combination of low cost, high latency (e.g. for storing ML training sets) and high cost, low latency systems (e.g. for real-time, integrated machine control feedback) may be needed.KeywordsFusionBig dataHigh performance computingData management

The White House has named acting leaders for the Office of Science and Technology Policy (OSTP) and the role of science adviser to the president. Current OSTP deputy director Alondra Nelson will serve as acting director. Former National Institutes of Health director Francis Collins will become acting science adviser and acting cochair of the President's Council of Advisors on Science and Technology (PCAST). Both Nelson and Collins will serve "until permanent leadership is nominated and confirmed," the White House says in a statement . This scientific reshuffle follows the resignation of Eric Lander as science adviser and OSTP director after a White House investigation found that he had created a hostile work environment . His last day was Feb. 18. "The selections are responsive to the dual importance of a strong OSTP that can drive science and technology solutions to our greatest challenges—and the very specific attention the President wants

At the close of each fiscal year, the US Department of Energy (DOE) Sustainability Performance Office (SPO) issues guidance documents and technical resource aids/tools necessary for DOE sites and national laboratories to complete sustainability reporting requirements. SPO is part of the DOE Office of Asset Management. As required by DOE Order 436.1A, Departmental Sustainability, each site develops and commits to an annual site sustainability plan (SSP) that identifies its respective contribution toward meeting DOE’s sustainability goals. SPO collects and compiles information reported by each site to develop an agency-wide sustainability report and implementation plan, which is used to report DOE sustainability progress to the federal government as required by all major federal agencies. DOE launched a formal Sustainability Office and annual SSP process in 2011. Each year, Oak Ridge National Laboratory (ORNL), in concert with the DOE Office of Science (SC), provides the resources essential to fulfill its commitment to deliver a complete and accurate SSP and quality performance data for entry into the DOE Sustainability Dashboard managed by SPO. The performance data entered by each DOE site are then combined to disclose the progress of each DOE program office and are further combined to show comprehensive progress for the agency. The Office of Asset Management assists program offices in sustaining their missions, freeing up resources by reducing waste, avoiding excess expenditure on utilities, maximizing productivity, and improving the efficiency of facilities and processes. By focusing on mission needs, programs and associated DOE sites can help the agency meet its sustainability goals as outlined in federal statutory and regulatory requirements. In FY 2022, the SSP Guidance/Instructions (US Department of Energy, Sustainability Performance Office) was updated to capture requirements from Executive Order (EO) 14008, Tackling the Climate Crisis at Home and Abroad; the Energy Act of 2020 (EAct 20); actions outlined in DOE’s Climate Adaptation and Resilience Plan and Sustainability Plan; and EO 14057, Catalyzing Clean Energy Industries and Jobs Through Federal Sustainability. Updates in SSP guidance help to minimize and streamline reporting while simultaneously addressing updated federal requirements. Per DOE, each SSP should provide an overview of the site’s planned actions and of efforts and accomplishments during the reporting period. SPO collects and compiles information reported by each site to develop DOE’s Annual Sustainability Report, Climate Adaptation and Resilience Plan, and Annual Energy Management Report to Congress. The agency goal has been to lower the reporting burden for sites and increase and improve the consistency of information available to decision-makers, allowing them to better identify projects and potential for increased efficiency and to reduce waste, lower emissions, and enhance operational resilience. Sites may elect to produce more polished publications for their leadership and stakeholders, but this step is no longer required. The ORNL SSP narrative report (this document) and the reporting of DOE SPO Sustainability Dashboard performance data is a collaborative effort of approximately 30 subject matter experts (SMEs) from ORNL facility management, mission support, and research divisions. Annually, these associates come together to provide a report that can be used by DOE to demonstrate continued agency progress in energy efficiency and sustainable federal operations.

The 21st Century will be the Information Age. Implementation and support of the strategic plan for a national/global information infrastructure will occur through the utilization of high-performance computing and communications (HPCC). HPCC technology provides a potentially huge payoff in health care. From teleconsultations to the availability of health-related databases, the effect of using HPCC will enhance patient care, improve drug design and broaden access to medical information. Geographic distance, time to accomplish tasks, separation of people from resources, and outdated organizational structures are impediments that inhibit the ultimate achievement of all these goals. Information technology has a pervasive and unprecedented ability to remove these barriers to progress. As part of the strategic planning process, three broad classes of user-driven applications have been identified: (1) high-performance applications; (2) high-confidence applications-improving the integration, privacy, security and reliability of information flows within and across enterprises; and (3) high-capability applications for the individual-empowering individuals with universal, easy-to-use access information, and providing customization and support of their information space.

The design, analysis and development of parallel and distributed applications in High Performance Computing and Communication (HPCC) Systems are still very challenging tasks. Therefore, there is a great need for an integrated multilevel analysis methodology to assist in designing and analyzing the performance of both existing and proposed new HPCC systems. Currently, there are no comprehensive analysis methods that address such hierarchical modeling approach for analyzing the end-to-end performance of an application running on an HPCC system. The overall system is partitioned into application level, protocol level and network level. Functions at each level are modeled using queueing networks. The approach enables the designer to study performance for different types of networks and protocols and different design strategies. In this paper we use video-on-demand as an application example to show how our approach can be used to analyze the performance of such an application.

The Federal HPCC Program is a 10-agency billion dollar per year effort to accelerate the development of high performance computing and communications technologies and the diffusion of these technologies to improve U.S. competitiveness and the well-being of citizens.

The US Department of Energy (DOE) Exascale Computing Project (ECP) has fostered and strengthened the use of modern software engineering practices for developing applications and libraries, and this effort has resulted in the coordinated and interoperable E4S1 and xSDK2 ecosystems. Although this approach is cost-effective, it relies on robust programming systems and tools (PST) as the underlying foundation for our HPC software. At present, our primary PST stack consists of traditional high-performance computing (HPC) languages, namely Fortran, C, C++, and the popular Python language for data analysis and AI workflows. These languages support various programming frameworks and run-time abstractions that enable parallelism and concurrency across multiple node architectures and thousands of nodes through a variety of interconnect systems. However, to accommodate users’ diverse needs, certain aspects of the HPC ecosystem are delegated to vendor-specific or third-party implementations that extend beyond a particular scientific domain. This broader scope results in a multitude of specifications and variations, which leads to a complex orchestration of many-ecosystems. Unfortunately, this complexity in the ecosystem imposes additional overhead costs on consumers during the latter stages of the development cycle. In addition to the software ecosystem challenge, the upcoming conclusion of the ECP by December 2023 has raised significant concerns within the HPC programming systems community, from both the economic and social perspectives. The ECP has implemented a management structure for software development and funding decisions across all ECP participants by following a conventional hierarchical and centralized approach. However, this structure has prompted certain considerations within the community, particularly in anticipation of the Software Sustainability initiative by the DOE’s Advanced Scientific Computing Research Program (ASCR). For the success of this new initiative, it is of utmost importance to secure consistent funding and foster close engagement with researchers and core developers of existing programming-system products. This collaboration is vital to maintaining the critical capabilities of the current software during the transition phase while proactively adapting to future technology and workforce trends. The community recognizes the significance of adapting to emerging trends and is aware of the inherent fragility of the HPC software ecosystem, particularly in relation to programming systems that cater to all users. The ability to adapt and evolve is essential to staying relevant and effectively addressing these technical, economic, and social challenges. The S4PST team, which represents one of the six ASCR Software Sustainability seedling projects, is dedicated to tackling these challenges through community-based approaches that go beyond the scope of the DOE. This involves collaboration between national laboratories with academia, non-DOE institutions, hardware and system vendors, and international partners. By fostering these partnerships, we aim to create a robust and sustainable HPC software ecosystem that can effectively meet the needs of the community. This new community effort, driven by the eight DOE labs, will take on the responsibility of guiding funding decisions for programming-systems development and maintenance with transparency and consistency across all decisions. Additionally, the team will offer common technical services to the programming systems community, irrespective of their funding situations, and facilitate community-wide incubation to proactively nurture the software ecosystem. By actively engaging with stakeholders and employing a collaborative approach, we can collectively shape the future of programming systems and ensure a robust and thriving HPC software landscape. On May 11–12, 2023, the S4PST team conducted its inaugural kick-off workshop at the Innovative Computing Laboratory (ICL) in the University of Tennessee, Knoxville, hosted by Hartwig Anzt. The workshop encompassed various sessions dedicated to presentations and discussions, with the aim of comprehending the team members’ perspectives on the vision of software sustainability. Additionally, the workshop aimed to identify the technical, economic, and social requirements for sustaining the programming-systems community in the field of HPC. This report provides a summary of the S4PST effort by highlighting five major thrust areas discussed during the workshop: (i) community, (ii) technical support, (iii) training and diversity, (iv) verification, validation and correctness, and (v) emerging technologies. It also encompasses an overview of the presentations and discussions held throughout the event, our views and potential synergies with other seedling efforts, along with the outcomes and key takeaways from our initial discussions.

This report documents the Tiger Team Assessment of the Los Alamos National Laboratory (LANL) located in Los Alamos, New Mexico. LANL is operated for the US Department of Energy (DOE) by the University of California. The Tiger Team Assessment was conducted from September 23 to November 8, 1991, under the auspices of the DOE Office of Special Projects, Office of Assistant Secretary for Environment, Safety and Health. The assessment was comprehensive, encompassing environmental, safety, and health (ES H) disciplines; management; and contractor and DOE self-assessments. Compliance with applicable Federal, state, and local regulations; applicable DOE Orders; best management practices; and internal LANL site requirements was assessed. In addition, an evaluation of the adequacy and effectiveness of the DOE and the site contractors' management of ES H/quality assurance programs was conducted. This volume discusses findings concerning the environmental assessment.

The Los Alamos Large Scale Demonstration and Deployment Project (LSDDP), in support of the US Department of Energy (DOE) Deactivation and Decommissioning Focus Area (DDFA), has been identifying and demonstrating technologies to reduce the cost and risk of management of transuranic element contaminated large metal objects, i.e. gloveboxes. DOE must dispose of hundreds of gloveboxes from Rocky Flats Environmental Technology Site (RFETS), Los Alamos National Laboratory (LANL), and other DOE sites. This paper reports on the results of four technology demonstrations on decontamination of plutonium-contaminated gloveboxes with each technology compared to a common baseline technology, wipedown with nitric acid. The general objective of the demonstrations was decontamination to an alpha-emitting nuclide disintegration rate of less than 50,000 disintegrations per minute per 100 square centimeters (dpm/100 cm2), the surface activity level desired for re-application of these particular gloveboxes to a new mission. The technologies demonstrated include: • A LANL-developed electrochemical decontamination system (EDS) technique utilizing a recycled electrolyte solution to contact the glovebox surface via a small electrode fixture, which is moved from location to location until the entire metal surface is decontaminated. • A commercial three-step decontamination technology marketed by Environmental Alternatives Inc. (EAI) was demonstrated to quantify its performance relative to the baseline technology. • Cerium (IV) nitrate decontamination, previously utilized at other DOE sites and developed for application to gloveboxes at RFETS, was demonstrated to quantify its performance in this application. • A Russian-developed electrochemical decontamination (ECD) technology was monitored by the Los Alamos LSDDP for potential application in DOE. Although this decontamination activity was not an LSDDP “demonstration,” it was planned, monitored, and reported using LSDDP methodologies. Generally, the experience from these demonstrations shows that all innovative technologies perform better than the baseline, nitric acid wipedown. The goal of meeting 50,000 dpm/100 cm2 was not achieved by the baseline technology or cerium nitrate decontamination at all measured locations with the number of decontamination technologies used in the demonstration. Additional decontamination cycles were estimated for achievement of the targeted activity for cost estimating purposes. However, the actual decontamination achieved may be acceptable for LLW status at some facilities. Both electrochemical techniques are capable of decontaminating surfaces to the targeted contamination level and, if desired, can decontaminate to very low levels. The EAI technology is the best performing of the wipedown techniques, but is more costly. Table I summarizes the number of cycles the various technologies required to achieve the desired decontamination level and the associated decontamination factor.

The United States Office of Science and Technology Policy (OSTP) delivered an open access directive in early 2013 mandating that federally funded research articles and associated data must be made accessible online to the public, free of charge. In response to this mandate, university and library organizations proposed the Shared Access Research Ecosystem (SHARE). With experience in the design and creation of large-scale federated repositories, bepress, best known for our hosted institutional repository platform Digital Commons, can offer unique insight into what makes a federated repository successful. This paper will outline the attributes of a successful shared national cross digital repository. At bepress, we have thought a lot about large shared digital repositories and have come up with a model for what we believe a U.S. national repository should include. Before we dive into the story, however, it is worth sharing some background about bepress to provide context for this paper. First, bepress was founded in 1999 by University of California Berkeley faculty members who were frustrated with the state of scholarly communications and wanted to improve it for their peers. Being founded by faculty, with one of the founders currently serving as our board chair, means that we approach everything we create with a faculty-centric focus. For example, Digital Commons was built to be a publishing platform as well as an institutional repository (IR). This allows libraries to provide a variety of services to faculty, including workflows for publishing books, journals, and conference proceedings, among others. Second, bepress is committed to trying new things. We strive to push the frontier of scholarly communications. We were, for example, the first to try the software-as-a-service model for institutional repository software. At the time, around the year 2002, the general thinking in the scholarly communications community was that repository software had to be built and managed locally. Now, with the advantages of the softwareas-a-service model more widely understood, all of the major IR platforms like DSPACE, Fedora, and Eprints have software solution providers offering IR services in the cloud. Third, we have been working with shared, aggregated repositories for many years. We built our first integrated, federated repository over ten years ago for the discipline of law. In November of 2012, we launched our latest iteration of a federated repository called the Digital Commons Network. The Digital Commons Network includes the work of more than 300 institutions and over 800,000 full-text objects, making it one of the largest databases of open access research in the world. While we had the ability to launch the Digital Commons Network earlier, it was crucially important to us to wait until the community of universities and colleges that use our platform had at least 500,000 articles in their repositories combined before deciding to aggregate our content. When we launched the Network in November of 2012, there was little discussion of anything similar on a national scale. That changed in February of 2013 after the United States Office of Science and Technology Policy (OSTP) issued an open access directive mandating that federally funded research articles and associated data must be made accessible online to the public, free of charge. According to one estimate by Heather Joseph (2013), Executive Director of the Scholarly Publishing and Academic Resources Coalition (SPARC), this could include as many as 200,000 articles a year across the government agencies. With each agency independently evaluating options and partners for compliance, the OSTP announcement kicked off a series of discussions among a wide ranging group of stakeholders,