Full System Optimization Research Articles

Enabling power measurement and control on Astra: The first petascale Arm supercomputer

SummaryAstra, deployed in 2018, was the first petascale supercomputer to utilize processors based on the ARM instruction set. The system was also the first under Sandia's Vanguard program which seeks to provide an evaluation vehicle for novel technologies that with refinement could be utilized in demanding, large‐scale HPC environments. In addition to ARM, several other important first‐of‐a‐kind developments were used in the machine, including new approaches to cooling the datacenter and machine. This article documents our experiences building a power measurement and control infrastructure for Astra. While this is often beyond the control of users today, the accurate measurement, cataloging, and evaluation of power, as our experiences show, is critical to the successful deployment of a large‐scale platform. While such systems exist in part for other architectures, Astra required new development to support the novel Marvell ThunderX2 processor used in compute nodes. In addition to documenting the measurement of power during system bring up and for subsequent on‐going routine use, we present results associated with controlling the power usage of the processor, an area which is becoming of progressively greater interest as data centers and supercomputing sites look to improve compute/energy efficiency and find additional sources for full system optimization.

Energy management strategy for a power-split hydraulic hybrid wheel loader

Energy management strategies for a power-split hydraulic hybrid wheel loader are studied in this paper. The differences between the powertrain and the energy management system for on-road vehicles and the powertrain and the energy management system for off-road vehicles are first identified. Unlike on-road vehicles where the engine powers only the drivetrain, the engine in a wheel loader powers both the drivetrain and the working hydraulic system. In a non-hybrid wheel loader, the two subsystems interfere with each other since they share the same engine shaft. By using a power-split powertrain, this not only allows for optimal engine operation and regenerative braking but also greatly reduces the interference between the drivetrain and the working functions. An energy management strategy based on dynamic programming is developed to give full system optimization including both the drivetrain and the working functions. Both a long loading cycle and a short loading cycle are studied in this paper. The dynamic-programming-based strategy is compared with a rule-based strategy using simulation studies.

Heat Transfer Enhancement in a High Aspect Ratio Channel with System Optimization

Outdoor digital information display (DID) installed near public areas has become a very popular medium of communication. Cooling performance of an outdoor DID system has become a critical issue from increasing system temperature due to heat generation at the back panel, in addition to irradiation from the sunlight. Gap distance between the heating plate and glass could control the main heat transfer rate considering its positive or negative effect on the pressure and temperature gradient at the wall. Additional reduction of maximum surface temperature can be achieved using tapered geometry with increased inlet gap distance, as well as guide vane installation at the system exit to lower pressure loss. Full system optimization has also been performed, including correlation for tapered geometry and guide vane effect. With system optimization, we can find the optimal gap clearance of parallel or tapered geometry for best performance.

Investigation of a hybrid renewable– microgeneration energy system for power and thermal generation with reduced emissions

A conceptual study is described into the hybridization of Stirling engine-based residential cogeneration systems with solar thermal systems. Simulation results of four hybrid system configurations applied in various locations in Canada are presented and compared to Base Case systems without solar input. Additional optimization cases are discussed. Adding solar collectors to a residential cogeneration system has a clear potential to reduce natural gas consumption and greenhouse gas emissions. The simulated cases showed a 10%–15% decrease in the consumption of natural gas, which corresponds to a greenhouse gas emission reduction of approximately 700–1200 kg/house/year (depending on configuration and location). Hybrid systems are complex and highly integrated systems. A full system optimization was therefore not possible in this study. Recommendations are given for further optimization of this type of systems.

Stability of antibody-conjugated gold nanoparticles in the endolysosomal nanoenvironment: implications for noninvasive radiofrequency-based cancer therapy

The use of noninvasive radiofrequency (RF) electric fields as an energy source for thermal activation of nanoparticles within cancer cells could be a valuable addition to the emerging field of nano-mediated cancer therapies. Based on investigations of cell death through hyperthermia, and offering the ability for total-body penetration by RF fields, this technique is thought to complement and possibly outperform existing nano-heat treatments that utilize alternative heat production via optical or magnetic stimuli. However, it remains a challenge to understand fully the complex RF-nanoparticle-intracellular interactions before full system optimization can be engineered. Herein we have shown that liver cancer cells can selectively internalize antibody-conjugated gold nanoparticles (AuNPs) through receptor-mediated endocytosis, with the nanoparticles predominantly accumulating and aggregating within cytoplasmic endolysosomes. After exposure to an external RF field, nonaggregated AuNPs absorbed and dissipated energy as heat, causing thermal damage to the targeted cancer cells. We also observed that RF absorption and heat dissipation is dependent on solubility of AuNPs in the colloid, which is pH dependent. Furthermore, by modulating endolysosomal pH it is possible to prevent intracellular AuNP aggregation and enhance thermal cytotoxicity in hepatocellular cancer cells. From the Clinical EditorGold nanoparticles absorb energy from RF fields and can exert hyperthermic effects leading to cell death. Combining this known effect with antibody-based targeting of the nanoparticles, selective cancer specific hyperthermia induced cell death therapies can be designed, as demonstrated in this article.

Flight test results for UAVs using boid guidance algorithms

A critical technology for operating groups of Uninhabited Aerial Vehicles (UAVs) is distributed guidance. Distributed guidance allows an operator to command several vehicles at the same time, reduces operator workload, and adds redundancy to the system. Some of the leading software candidates for achieving distributed guidance are known as Boid Guidance Algorithms (BGAs), which are agent-based techniques relying on the interactions of simple behaviors. Flight tests are crucial to the advancement of flight technologies such as BGAs, and this was identified as an important area for development. This paper presents the results from the 2005 flight tests of BGAs at NASA Dryden Flight Research Center with two RnR Products’ APV-3 UAVs employing CloudCap Technology's Piccolo autopilot system. Major challenges in these flight tests include the use of a waypoint-following system, limited computation resources, and management of safety procedures. The conclusions of this work include the need for using a path-following platform and completion of a full system optimization. This work is an important step in the development of a deployable distributed guidance system.

Optimization and integration of high‐performance ground penetrating imaging radar system: A research prototype

AbstractThis paper reports the design, development, and field experimentation of a new ground penetrating imaging radar system. This research prototype unit features full system optimization, which successfully integrates the key components of antenna sensitivity, data‐acquisition waveforms, synthetic‐aperture scan, and image reconstruction algorithms, for optimal system performance.© 2006 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 15, 220–223, 2005