Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Resource Allocation”

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 865 records · Page 48
All sourcesNASA NTRSDOE OSTI
NASA NTRS1987

Cross-modal interference and task integration - Resources or preemption switching?

Data are reviewed from experiments that have contrasted intra-modal (visual-visual) information presentation with cross-modal (visual-auditory) presentation. Five different processing mechanisms that are operating in dual stimulus tasks are described, and it is concluded that in studies where visual scanning is not required, cross-modal effects are of two classes. When the visual task is continuous (tracking), a discrete auditory stimulus will preempt tracking performance relative to a discrete visual stimulus, leading to an effective shift in allocation bias. When both tasks are discrete, the data regarding the relative advantages of cross- vs. intra-modal interference are ambivalent.

Wickens, Christopher D.↗
NASA NTRS2018

Impact of Different Trajectory Option Set Participation Levels within an Air Traffic Management Collaborative Trajectory Option Program

This paper presents the methodology and results of a Human-In-The-Loop (HITL) simulation study conducted in the Airspace Operations Laboratory at NASA Ames Research Center. This study is a part of NASA's ongoing research into developing an Integrated Demand Management (IDM) concept, whose aim is to improve traffic flow management (TFM) by coordinating the FAA's strategic Traffic Flow Management System (TFMS) with its more tactical Time-Based Flow Management (TBFM) system. The purpose of TFM is to regulate air traffic demand so that it is delivered efficiently through constrained airspace resources without exceeding their capacity limits. The IDM concept leverages a new TFMS capability called the Collaborative Trajectory Options Program (CTOP) to strategically pre-condition traffic demand flowing into a TBFM-managed arrival environment, where TBFM is responsible for managing traffic tactically by generating precise arrival schedules. Unlike other TFM tools, CTOP gives flight operators the option of submitting a set of user-preferred alternative trajectories for each flight. CTOP can then use these trajectory option sets (or TOSs) to find user-preferred alternative routes to reduce demand on an overloaded resource. CTOP's effectiveness in redistributing demand is limited, however, by the availability of flights with alternative routes. The research presented in this paper focuses on evaluating the impact on TFM operations by varying the percentage of flights that submit a multiple-option TOS ('TOS participation levels'). Results show the impact on overall system performance and on the rerouted flights themselves. The simulation used a Newark (EWR) airport arrival scenario, with en route weather affecting traffic inbound from the west. Participants were asked to control each of the three arrival flows (north, west, and south) to meet their individual capacity constraints while simultaneously ensuring efficient utilization of the capacity at the destination airport. A large, permeable convective weather cell located southeast of Chicago severely reduced the capacity of the west flow. The study evaluated the impact of five different TOS participation levels on CTOP's ability to re-allocate traffic from the west and improve TFM performance in terms of delay assignment and traffic delivery rate to the airport. Overall, the results showed that increasing TOS submissions allowed the overall system delays to be reduced and fairly distributed among the three arrival flows, at the same time achieving the airport throughput rate. Moreover, it was found that aircraft who submitted a TOS saw a greater reduction in delay, even when they were assigned longer routes. This was particularly true when fewer aircraft submitted a TOS. The results confirm that the CTOP operations with higher TOS participation levels helped utilize the overall National Airspace System (NAS) resources as well as benefited the users who participated.

Yoo, Hyo-Sang↗
NASA NTRS1989

Ka-band study: 1988

The Ka-band study team was chartered in late 1987 to bring together all the planning elements for establishing 32 GHz (Ka-band) as the primary downlink frequency for deep-space operation, and to provide a stable baseline from which to pursue that development. This article summarizes the results of that study at its conclusion in mid-1988, and corresponds to material presented to NASA's Office of Space Operations on July 14, 1988. For a variety of reasons, Ka-band is the right next major step in deep-space communications. It offers improved radio metric accuracy through reduced plasma sensitivity and increased bandwidth. Because of these improvements, it offers the opportunity to reduce costs in the flight radio system or in the DSN by allocating part of the overall benefits of Ka-band to this cost reduction. A mission scenario is being planned that can drive at least two and possibly all three of the DSN subnets to provide a Ka-band downlink capability by the turn of the century. The implementation scenario devised by the study team is believed to be feasible within reasonable resource expectations, and capable of providing the needed upgrade as a natural follow-on to the technology development which is already underway.

Layland, J. W.↗
DOE OSTI2023

The Effect of Prosumer Duality on Power Market: The Effect of Market Regulation

Electricity prosumers are energy subsystems that not only consume, but also produce electricity. They are present in distribution level networks as traditional utility customers who have installed distributed energy resources. They are also present in transmission level networks as large conglomerates own both generation assets and large industrial loads. Previous work on the economics of prosumers has demonstrated that prosumers in a market have incentives to behave more competitively compared to producers and consumers in traditional markets. This paper further explores the behavior of prosumers and their response to market policies including the allocation of network losses and the impact of net metering. We extend a Cournot model of a dual prosumer and find that prosumers respond with higher supply quantities if network losses are allocated to demand base. They respond with lower supply quantities if network losses are allocated to supply base. Allocating network losses to demand base also causes equilibrium prices to decline. Markets where prosumers first satisfy their own load and then sell the balance of electricity to the grid have lower quantities and higher prices compared to markets where prosumers buy and sell at locational marginal price.

Tsybina, Eve↗
DOE OSTI2020

Probabilistic Zonal Reserve Requirements for Improved Energy Deliverability with Wind Power

In power systems with high penetration of renewable energy resources, uncertainty and variability of these stochastic resources introduce additional challenges for the operation of power systems. To improve the power system's reliability in the face of uncertainty, reserves are required as additional generation capacity to rebalance the power system following random disturbances. However, reserve deliverability is not guaranteed, because it may encounter potential transmission line congestion. Zonal reserve requirements can address this issue, but operators lack efficient ways to allocate reserves to zones while accounting for wind power forecast uncertainty. Herein we propose a methodology for probabilistic zonal reserve requirements to address wind power forecast uncertainties. This method estimates the probability distribution of line flows based on the system generation margin and injection shift factor. This estimate is then used to construct pre-defined and post-zonal reserve requirements. Case studies demonstrate that the proposed method efficiently schedules energy and reserves to balance energy and manage deliverability with wind power forecast uncertainty.We also discuss operational implications of the proposed method.

17 WIND ENERGY↗
NASA NTRS1992

Multi-agent planning and scheduling, execution monitoring and incremental rescheduling: Application to motorway traffic

This article describes a planning method applicable to agents with great perception and decision-making capabilities and the ability to communicate with other agents. Each agent has a task to fulfill allowing for the actions of other agents in its vicinity. Certain simultaneous actions may cause conflicts because they require the same resource. The agent plans each of its actions and simultaneously transmits these to its neighbors. In a similar way, it receives plans from the other agents and must take account of these plans. The planning method allows us to build a distributed scheduling system. Here, these agents are robot vehicles on a highway communicating by radio. In this environment, conflicts between agents concern the allocation of space in time and are connected with the inertia of the vehicles. Each vehicle made a temporal, spatial, and situated reasoning in order to drive without collision. The flexibility and reactivity of the method presented here allows the agent to generate its plan based on assumptions concerning the other agents and then check these assumptions progressively as plans are received from the other agents. A multi-agent execution monitoring of these plans can be done, using data generated during planning and the multi-agent decision-making algorithm described here. A selective backtrack allows us to perform incremental rescheduling.

Mourou, Pascal↗
NASA NTRS1991

Using C to build a satellite scheduling expert system: Examples from the Explorer Platform planning system

A C-based artificial intelligence (AI) development effort which is based on a software tools approach is discussed with emphasis on reusability and maintainability of code. The discussion starts with simple examples of how list processing can easily be implemented in C and then proceeds to the implementations of frames and objects which use dynamic memory allocation. The implementation of procedures which use depth first search, constraint propagation, context switching, and blackboard-like simulation environment are described. Techniques for managing the complexity of C-based AI software are noted, especially the object-oriented techniques of data encapsulation and incremental development. Finally, all these concepts are put together by describing the components of planning software called the Planning And Resource Reasoning (PARR) Shell. This shell was successfully utilized for scheduling services of the Tracking and Data Relay Satellite System for the Earth Radiation Budget Satellite since May of 1987 and will be used for operations scheduling of the Explorer Platform in Nov. of 1991.

Mclean, David R.↗
NASA NTRS1991

Using C to build a satellite scheduling expert system: Examples from the Explorer platform planning system

Recently, many expert systems were developed in a LISP environment and then ported to the real world C environment before the final system is delivered. This situation may require that the entire system be completely rewritten in C and may actually result in a system which is put together as quickly as possible with little regard for maintainability and further evolution. With the introduction of high performance UNIX and X-windows based workstations, a great deal of the advantages of developing a first system in the LISP environment have become questionable. A C-based AI development effort is described which is based on a software tools approach with emphasis on reusability and maintainability of code. The discussion starts with simple examples of how list processing can easily be implemented in C and then proceeds to the implementations of frames and objects which use dynamic memory allocation. The implementation of procedures which use depth first search, constraint propagation, context switching and a blackboard-like simulation environment are described. Techniques for managing the complexity of C-based AI software are noted, especially the object-oriented techniques of data encapsulation and incremental development. Finally, all these concepts are put together by describing the components of planning software called the Planning And Resource Reasoning (PARR) shell. This shell was successfully utilized for scheduling services of the Tracking and Data Relay Satellite System for the Earth Radiation Budget Satellite since May 1987 and will be used for operations scheduling of the Explorer Platform in November 1991.

Mclean, David R.↗
DOE OSTI2026

Distributed optimization and control for autonomous distributed energy resource power dispatch and frequency regulation considering communication failures

Distributed energy resources (DERs) can significantly enhance grid flexibility by supplying power and secondary frequency regulation (SFR) services. Yet, their large number and dispersed nature limit the effectiveness of traditional centralized power dispatch and frequency control schemes, and expose the system to single-point communication failures. This paper presents a unified distributed optimization and control framework for DER power dispatch and SFR provision. A transmission-loss-aware distributed direct current optimal power flow is first formulated to optimize DER power dispatch and reserve scheduling. On top of this, a fully distributed control scheme for DER autonomous SFR provision is developed to allocate automatic generation control signals cost-effectively while ensuring robustness to communication link disruptions. The framework is validated on IEEE 14-bus and 39-bus transmission systems with large-scale distribution feeders using transmission-distribution dynamic co-simulation, demonstrating reliable and autonomous DER power dispatch and SFR provision under diverse operating conditions and communication contingencies.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRSDate not supplied

CHP-PRA: Sensorimotor Countermeasures Proof of Concept

The capabilities included in Crew Health and Performance (CHP) systems are designed to keep the crew healthy, happy, and productive. In doing so, the CHP system allows reductions in human related risks to be realized. All human missions, though particularly future Artemis and Mars missions, have constraints on the mass and volume allocated to the CHP system. Thus, trades must be made on how to best buy down risk while still meeting other requirements. Understanding how different CHP capabilities influence medical, performance, and long-term health risks is key to making informed choices among these trades. To address this, an integrated CHP Probabilistic Risk Assessment (PRA) model is being developed. Much like how IMPACT is designed to allow medical resource trades informed by medical risks, CHP PRA will enable analogous trades in human system risks across all CHP functions and capabilities. There are 29 Human System Risk Board (HSRB) risk areas to consider for model implementation. This effort focused solely on the implementation of countermeasures associated with sensorimotor risk. The framework and lessons learned from this proof of concept will benefit the implementation of other risks and countermeasures. Astronauts experience sensorimotor changes when entering or exiting a microgravity environment. While most sensorimotor issues are resolved within a few days, they pose a serious risk to crew health and performance during the adaptation period. This adaptation period aligns with gravity transitions and key mission phases where the crew members may be required to complete challenging tasks ideally with an undisturbed sensorimotor system. Countermeasures such as training or pharmaceuticals can be used to mitigate risk. This effort began by identifying physiological changes to the sensorimotor system that could lead to functional performance decrements or medical conditions and potential countermeasures, so that the overall effect of sensorimotor disturbance on human risk could be captured. We then selected example countermeasures, mapped relationships between countermeasures and outcomes, and used results from existing sensorimotor investigations to define the relationship between a sensorimotor countermeasure and subsequent performance of a sensorimotor-affected task. Specifically, using shuttle landing performance data, we compared the use or absence of inflight training capabilities to yield a relative performance change that can be applied to the prediction of the risk associated with manual control performance in future missions, as depicted in the figure. This sensorimotor proof of concept demonstrates how HRP research information can be transformed into quantifiable relationships to describe changes in HSRB medical and performance risks.

Caroline R Austin↗
DOE OSTI2022

North Carolina Water Utility Builds Resilience with Distributed Energy Resources

As the frequency and duration of grid outages increase, backup power systems are becoming more important for ensuring that critical infrastructure continues to provide essential services. Most facilities rely on diesel generators, which may be ineffective during long outages owing to limited fuel supplies and high generator failure rates. Distributed energy resources such as solar, storage, and combined-heat-and-power systems, coupled with on-site biofuel production, offer an alternative source of on-site generation that can provide both cost savings and resilience (i.e., the ability to respond to catastrophic events with longer-term consequences). A mixed-integer linear program minimizes costs and maximizes resilience at a wastewater treatment plant in Wilmington, North Carolina. We find that the plant can reduce life-cycle energy costs by 3.1% through the installation of a hybrid combined-heat-and-power, photovoltaic, and storage system. When paired with existing diesel generators, this system can sustain full load for seven days while saving $664,000 over 25 years and reducing diesel fuel use by 48% compared with the diesel-only solution. This analysis informed a decision by the Cape Fear Public Utility Authority to allocate funds for the implementation of a combined-heat-and-power system at the wastewater treatment plant in fiscal year 2023. Finally, the benefits of deploying hybrid combined-heat-and-power technologies and the utilization of on-site biofuel production extend, on a national scale, to thousands of wastewater treatment facilities and other types of critical infrastructure.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2023

Modeling the Water Systems of the Western US to Support Climate‐Resilient Electricity System Planning

Abstract Electricity and water systems in the Western US (WUS) are closely connected, with hydropower comprising 20% of total annual WUS generation, and electricity related to water comprising about 7% of total WUS electricity use. Because of these interdependencies, the threat of climate change to WUS resources will likely have compounding electricity impacts on the Western Interconnect grid. This study describes a WUS‐wide water system model with a particular emphasis on estimating climate impacts on hydropower generation and water‐related electricity use, which can be linked with a grid expansion model to support climate‐resilient electricity planning. The water system model combines climatically‐driven physical hydrology and management of both water supply and demand allocation, and is applied to an ensemble of 15 climate scenarios out to 2050. Model results show decreasing streamflow in key basins of the WUS under most scenarios. Annual water‐related electricity use increases up to 4%, and by up to 6% during the summer months, driven by growing agricultural demands met increasingly through a shift toward energy‐intensive groundwater to replace declining surface water. Total annual hydropower generation changes by +5% to −20% by mid‐century but declines in most scenarios, with decreases in summer generation by up to nearly −30%. Water‐related electricity use increases tend to coincide with hydropower generation declines, annually and seasonally, demonstrating the importance of concurrently evaluating the climate signal on both water‐for‐energy and energy‐for‐water to inform planning for grid reliability and decarbonization goals.

13 HYDRO ENERGY↗
NASA NTRS1981

Spacecraft computer resource margin management

The conduction of the Project Galileo Orbiter, with 18 microcomputers and the equivalent of 360K 8-bit bytes of memory contained within two major engineering subsystems and eight science instruments, requires that the key onboard computer system resources be managed in a very rigorous manner. Attention is given to the rationale behind the project policy, the development stage, the preliminary design stage, the design/implementation stage, and the optimization or 'scrubbing' stage. The implementation of the policy is discussed, taking into account the development of the Attitude and Articulation Control Subsystem (AACS) and the Command and Data Subsystem (CDS), the reporting of margin status, and the response to allocation oversubscription.

Larman, B. T.↗
DOE OSTI2020

US Department of Energy, Office of Science High Performance Computing Facility Operational Assessment 2019 Oak Ridge Leadership Computing Facility

Oak Ridge National Laboratory's (ORNL's) Leadership Computing Facility (OLCF) continues to surpass its operational target goals: supporting users; delivering fast, reliable computational ecosystems; creating innovative solutions for high performance computing (HPC) needs; and managing risks, safety, and security associated with operating some of the most powerful computers in the world. The results can be seen in the cutting-edge science conducted by users and the praise from the research community. Calendar year (CY) 2019 was a big year as OLCF staff ran five world-class resources (the leadershipclass computers Titan and Summit, the large analysis cluster called Eos, and the massive parallel filesystems called Atlas and Alpine)) and also began power and cooling upgrades for a 2021 exascale system called Frontier. While continuing exceptional operation of Titan, Eos, and Rhea, the OLCF released the Summit supercomputer for production on January 1, 2019. Summit debuted as the most capable and efficient system in its class and has been recognized as the most powerful system in the world for its performance on both the high performance linpack (HPL) and conjugate gradient (HPCG) benchmark applications since June 2018 according to TOP500. Summit represents the culmination of a multiyear effort between the OLCF, IBM, NVIDIA, and Mellanox to deliver a system that is unmatched for modeling, simulation, data analysis, and learning. To hit the ground running with science-ready applications on day one, application teams worked closely with the OLCF through the Center for Accelerated Application Readiness (CAAR) program for years in advance of the Summit deployment. CY 2019 was filled with outstanding results and accomplishments: a very high rating from users on overall satisfaction for the sixth year in a row; a tremendous amount of core-hours delivered to researchers from two leadership-class systems; and success in delivering on the allocation split of roughly 60%, 30%, and 10% of core-hours offered for the Innovative and Novel Computational Impact on Theory and Experiment (INCITE), Advanced Scientific Computing Research Leadership Computing Challenge (ALCC), and Director's Discretionary (DD) programs, respectively (see Operational Performance section). These accomplishments, coupled with the high utilization rates (overall and capability usage), represent the fulfillment of the promise of both leadership-class machines: efficient facilitation of leadership-class computational applications. Table ES.1 presents a summary of the 2019 OLCF metric targets and the associated results. More information can be found in the Operational Performance section for each OLCF resource. The scientific accomplishments of OLCF users are a strong indication of long-term operational success, with publications this year in such notable journals and publications as Nature, Nature Physics, Nature Plants, Physical Review X, Journal of the American Physical Society, Cell, Nano Letters, and Trends in Biotechnology. Crucial domain-specific discoveries facilitated by resources at the OLCF are described in the High Performance Computing Facility Operational Assessment 2019 Oak Ridge Leadership Computing Facility (OAR) Strategic Results section. For example, researchers used Summit to pinpoint and understand the production of proteins from genetic information, including mutations and the functional expression of disease (Section 8.2).

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2018

From Apollo to the Future, the NASA Curation Model for Engaging the Sample Science Community Maximizes Science on Extraterrestrial Samples

The Astromaterials Acquisition and Curation Office at Johnson Space Center (JSC) has enjoyed a long-term partnership (50 years!) with a broad community of planetary sample scientists. This partnership has enabled the curators of planetary samples to plan for and enact evolving requirements for preservation of sample scientific integrity and for handling and long-term storage. The basis for this relationship is a standing peer review advisory committee composed of leading scientists who are recognized for achievements in sample analysis. The committee and its descendants have brought familiarity with the most relevant scientific investigations and the associated analytical and contamination challenges. Beginning with Apollo, the review committee was charged with oversight of curatorial operations and with ensuring fair access to samples. As additional samples from other planetary bodies were acquired, the committee evolved, taking on new responsibilities, reflected in committee name changes. However, oversight of curatorial operations and fair allocation of samples remain basic responsibilities. Committee recommendations are sent to the NASA Headquarters Discipline Scientist for approval. To minimize conflict of interest and maximize fair access, the rules governing the make-up of the committee is structured. Systematic rotation of leadership and staggered terms of membership allow the committee to retain expertise while bringing in fresh ideas. The first peer review committee was called the Lunar Sample Analysis and Planning Team (LSAPT) and was formalized in early 1968 with about 15 members. Their function was to review a) the equipment and procedures used in the new Lunar Receiving Laboratory (LRL); b) the proficiency and capability of the LRL staff; c) the sequence of sample analysis and allocation after quarantine release; and d) the findings of the Preliminary Examination Team (PET). According to LSAPT member Gerald Wasserburg, one of the first issues they faced was deciding whether to have most of the sample analyses performed in house at the LRL or to distribute samples to members of the scientific community. LSAPT concluded that the major scientific investigations should be carried out externally to the LRL by scientists chosen for their expertise in specific disciplines. Further they recommended that the PET's basic characterization of samples be circulated to the broad scientific community. LSAPT set its own agenda, paid attention to facility details, closely monitored the move of samples from the LRL to the interim curatorial facility in 1973, and was active in inspecting curation facilities. Between 1975 and 1979, a Facility Subcommittee of LSAPT oversaw the design and construction of a permanent facility for preservation of lunar samples. The result was an outstanding facility still in use today. In 1977, a separate peer review committee, the Meteorite Working Group (MWG), was formed to evaluate requests for new meteorites then being collected in Antarctica under what would in 1980 become a 3-agency agreement (National Science Foundation, NASA, Smithsonian Institution). By 1979, after lunar samples were moved into the new permanent facility, the vacated gloveboxes and laboratory were prepared for meteorite curation. Recognizing that LSAPT had been helpful in setting up the JSC curatorial facility for Antarctic meteorites, JSC recommended the review committee be given expanded duties, including advice on curation and analysis of materials from other planetary bodies and the name be changed to Lunar and Planetary Sample Team (LAPST). In 1993, LAPST was renamed the Curation and Analysis Planning Team for Extraterrestrial Materials (CAPTEM) to reflect additional functions. CAPTEM is chartered to be (1) a community-based, interdisciplinary forum for discussion and analysis of matters concerning the collection and curation of extraterrestrial samples, including planning future sample return missions and (2) a standing review panel, charged with evaluating proposals requesting allocation of all extraterrestrial samples contained in NASA collections. Efficiency and flexibility are gained through use of subcommittees, both ad hoc and standing. Transition of the MWG to a subcommittee of CAPTEM was completed in 2017. Today subcommittees review allocation requests for lunar samples, Antarctic meteorites, cosmic dust, Stardust cometary samples, Genesis solar wind samples, and samples returned from asteroids. Other subcommittees address facilities, informatics, and micro-cratered substrates. Planetary samples have been sent to research teams in over 30 countries world-wide. The expertise in the care and fair distribution of astromaterials by NASA using this model spans generations of planetary sample scientists and is a valuable resource to be tapped for future sample returns - OSIRIS-REx, Hayabusa 2, and Mars 2020.

Allton, Judith↗
DOE OSTI2020

SQUARE: Strategic Quantum Ancilla Reuse for Modular Quantum Programs via Cost-Effective Uncomputation

Compiling high-level quantum programs to machines that are size constrained (i.e. limited number of quantum bits) and time constrained (i.e. limited number of quantum operations) is challenging. In this paper, we present SQUARE (Strategic QUantum Ancilla REuse), a compilation infrastructure that tackles allocation and reclamation of scratch qubits (called ancilla) in modular quantum programs. At its core, SQUARE strategically performs uncomputation to create opportunities for qubit reuse. Current Noisy Intermediate-Scale Quantum (NISQ) computers and forward-looking Fault-Tolerant (FT) quantum computers have fundamentally different constraints such as data locality, instruction parallelism, and communication overhead. Our heuristic-based ancilla-reuse algorithm balances these considerations and fits computations into resource-constrained NISQ or FT quantum machines, throttling parallelism when necessary. To precisely capture the workload of a program, we propose an improved metric, the "active quantum volume," and use this metric to evaluate the effectiveness of our algorithm. Furthermore, our results show that SQUARE improves the average success rate of NISQ applications by 1.47X. Surprisingly, the additional gates for uncomputation create ancilla with better locality, and result in substantially fewer swap gates and less gate noise overall. SQUARE also achieves an average reduction of 1.5X (and up to 9.6X) in active quantum volume for FT machines.

compiler optimization↗
DOE OSTI2025

A Portfolio Approach to Massively Parallel Bayesian Optimization

One way to reduce the time of conducting optimization studies is to evaluate designs in parallel rather than just one-at-a-time. For expensive-to-evaluate black-boxes, batch versions of Bayesian optimization have been proposed. They work by building a surrogate model of the black-box to simultaneously select multiple designs via an infill criterion. Still, despite the increased availability of computing resources that enable large-scale parallelism, the strategies that work for selecting a few tens of parallel designs for evaluations become limiting due to the complexity of selecting more designs. It is even more crucial when the black-box is noisy, necessitating more evaluations as well as repeating experiments. Here we propose a scalable strategy that can keep up with massive batching natively, focused on the exploration/exploitation trade-off and a portfolio allocation. We compare the approach with related methods on noisy functions, for mono and multi-objective optimization tasks. These experiments show orders of magnitude speed improvements over existing methods with similar or better performance.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1998

The SGI/Cray T3E: Experiences and Insights

The NASA Goddard Space Flight Center is home to the fifth most powerful supercomputer in the world, a 1024 processor SGI/Cray T3E-600. The original 512 processor system was placed at Goddard in March, 1997 as part of a cooperative agreement between the High Performance Computing and Communications Program's Earth and Space Sciences Project (ESS) and SGI/Cray Research. The goal of this system is to facilitate achievement of the Project milestones of 10, 50 and 100 GFLOPS sustained performance on selected Earth and space science application codes. The additional 512 processors were purchased in March, 1998 by the NASA Earth Science Enterprise for the NASA Seasonal to Interannual Prediction Project (NSIPP). These two "halves" still operate as a single system, and must satisfy the unique requirements of both aforementioned groups, as well as guest researchers from the Earth, space, microgravity, manned space flight and aeronautics communities. Few large scalable parallel systems are configured for capability computing, so models are hard to find. This unique environment has created a challenging system administration task, and has yielded some insights into the supercomputing needs of the various NASA Enterprises, as well as insights into the strengths and weaknesses of the T3E architecture and software. The T3E is a distributed memory system in which the processing elements (PE's) are connected by a low latency, high bandwidth bidirectional 3-D torus. Due to the focus on high speed communication between PE's, the T3E requires PE's to be allocated contiguously per job. Further, jobs will only execute on the user specified number of PE's and PE timesharing is possible but impractical. With a highly varied job mix in both size and runtime of jobs, the resulting scenario is PE fragmentation and an inability to achieve near 100% utilization. SGI/Cray has provided several scheduling and configuration tools to minimize the impact of fragmentation. These tools include PScheD (the political scheduler), GRM (the global resource manager) and NQE (the Network Queuing Environment). Features and impact of these tools will be discussed, as will resulting performance and utilization data. As a distributed memory system, the T3E is designed to be programmed through explicit message passing. Consequently, certain assumptions related to code design are made by the operating system (UNICOS/mk) and its scheduling tools. With the exception of HPF, which does run on the T3E, however poorly, alternative programming styles have the potential to impact the T3E in unexpected and undesirable ways. Several examples will be presented (preceeded with the disclaimer, "Don't try this at home! Violators will be prosecuted!")

Bernard, Lisa Hamet↗