Search NASA⌕ Search

Engineering topics

Patil, Deepak

Publications and source records attributed to Patil, Deepak.

ICED: An Integrated CGRA Framework Enabling DFVS-Aware Acceleration

oarse-grained reconfigurable arrays (CGRAs) are a promising solution to enable energy-efficient acceleration of applications from different domains. By leveraging reconfiguration at the functional level, they can adapt to significantly different computational patterns. Existing CGRA mapping approaches extract instruction-level parallelism, exploit loop-pipelining opportunities, guarantee the data dependency, and target high throughput of a given loop. However, the recurrence data-dependency in the DFG and the mismatch between required and available computing/communication resources complicate the mapping, and might lead to significant unbalances in the utilization of the CGRA's tiles. This results in wasted power for tiles with low utilization. Applying dynamic voltage and frequency scaling (DVFS) can potentially solve this challenge and improve energy efficiency by adjusting voltage and frequency of different tiles independently. CGRAs have also been successful in accelerating data-dependent streaming applications. However, in these applications, the execution time of each kernel in the pipeline might dynamically vary depending on the characteristics of the input. This also leads to under-utilization of resources for the dynamically changing kernels that do not limit the application throughput. DVFS can also improve energy efficiency for these applications by dynamically changing the voltage and frequency levels of tiles that host non performance-constraining kernels. This paper proposes ICEDTEA -- an integrated DVFS-aware framework to map applications on CGRAs that support power islands. ICEDTEA proposes a CGRA architecture supporting DVFS islands at varying granularity (from a single tile to a group of tiles) and the related DVFS-aware compilation and mapping toolchain. ICEDTEA is the first work that introduces DVFS support for spatio-temporal CGRAs at power-island levels. The experimental evaluation shows that ICEDTEA improves average utilization by 2.3$\times$ and energy-efficiency by 1.32$\times$ over a conventional CGRA. With streaming applications, ICEDTEA improves energy efficiency by 1.12$\times$ over a state-of-the-art CGRA that introduces partial dynamic reconfiguration to adapt to variations in kernels' throughput.

Tan, Cheng↗

VecPAC: A Vectorizable and Precision-Aware CGRA

This paper proposes VecPAC -- a vectorizable and precision-aware coarse-grained reconfigurable array (CGRA) design. VecPAC integrates CGRA tiles with scalar functional units and specialized tiles with vector functional units that can trade off the number of vector lanes for the accuracy of the computation. We discuss the architecture design and present the related compilation framework. The experimental evaluation on a set of applications from three different domains (embedded, machine learning, and high-performance computing) shows that the hybrid design of VecPAC outperforms CGRAs with only scalar functional units by 1.48x, while providing higher scalability.

Tan, Cheng↗

Flux crystal growth of uranium(v) containing oxyfluoride perovskites

The novel phases Rb 4 NaU 3 O 12-x F x (1), K 4 NaU 3 O 12-x F x (2), and Rb 2.1 K 1.9 KU 3 O 12-x F x (3) were synthesized by molten flux methods using mixed alkali fluoride melts. The oxyfluorides crystallize in the cubic space group Im[3 with combining macron]m with a lattice parameters of 8.7472(2) Å, 8.6264(2) Å, and 8.8390(3) Å, respectively. All three structures crystallize in a cubic perovskite structure, ABO 3 (A 4 BB' 3 O 12 ), where the A site is fully occupied by an alkali cation, and the B site is shared by the remaining smaller alkali cation and uranium in an ordered fashion such that the alkali cation on the B site is surrounded by square uranyl bipyramids. The structures were characterized by single crystal X-ray diffraction, energy dispersive spectroscopy, X-ray absorption near edge structure spectroscopy, X-ray photoelectron spectroscopy, magnetic susceptibility measurements, DFT calculations, thermogravimetric analysis, and UV-vis spectroscopy, all of which support the presence of U(V) in the three new materials.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Glass structure and crystallization in boro-alumino-silicate glasses containing rare earth and transition metal cations: a US-UK collaborative program

Nuclear wastes generated from reprocessing of used nuclear fuel tend to contain a large fraction of rare earth (RE, e.g., Nd3+), transition (TM, e.g., Mo6+, Zr4+), alkali (A, e.g., Cs+), and alkaline earth cations (AE, e.g., Ba2+, Sr2+). Various strategies have been considered for immobilizing such waste streams, varying from nominally crystal-free glass to glass-ceramic to multi-phase ceramic waste forms. For glass and glass-ceramic waste forms, the added glass-forming system is generally alkali-alkaline earth-aluminoborosilicate (i.e., Na-Ca-Al-B-Si oxide). In a US-UK collaborative project, summarized here, we investigated the glass structure and crystallization dependence on compositional changes in simulated nuclear waste glasses and glass-ceramics. Compositions ranged in complexity from five – to – eight oxides. Specifically, the roles of Mo and rare earths are investigated, since a proposed glass-ceramic waste form contains crystalline phases such as powellite [(AE,A,RE)MoO4] and oxyapatite [(RE,AE,A)10Si6O26], and the precipitation of molybdenum phases is known to be affected by the rare earth concentration in the glass. Additionally, the effects of other chemical additions have been systematically investigated, including Zr, Ru, P, and Ti. A series of studies were also undertaken to ascertain the effect of the RE size on glass structure and on partitioning to crystal phases, investigating similarities and differences in glasses containing single RE oxides of Sc, Y, La, Ce, Nd, Sm, Er, Yb, or Lu. Finally, the effect of charge compensation was investigated by considering not only the commonly assessed peralkaline glass but also metaluminous and peraluminous compositions. Glass structure and crystallization studies were conducted by spectroscopic methods (i.e., Raman, X-ray absorption, nuclear magnetic resonance (NMR), optical absorption, photoluminescence, photoluminescence excitation, X-ray photoelectron spectroscopy), microscopy (i.e., scanning electron microscopy, transmission electron microscopy, electron probe microanalysis), scattering (i.e., X-ray and neutron diffraction, small angle measurements), and physical characterization (i.e., differential thermal analysis, liquidus, viscosity, density). This paper will give an overview of the research program and some example unpublished results on glass-ceramic crystallization kinetics, microstructure, and Raman spectra, as well as some examples of the effects of rare earths on the absorption, luminescence, and NMR spectra of starting glasses. The formal collaboration described here has resulted in the generation of a large number of results, some of which are still in the process of being published as separate studies.

McCloy, John S.↗

Method Development for High Temperature In-Situ Neutron Diffraction Measurements of Glass Crystallization on Cooling from Melt

A glass-ceramic borosilicate waste form is being considered for immobilization of waste streams of alkali, alkaline-earth, lanthanide, and transition metals generated by transuranic extraction for reprocessing used nuclear fuel. Waste forms are created by partial crystallization on cooling, primarily of oxyapatite and powellite phases. In-situ neutron diffraction experiments were performed to obtain detailed information about crystallization upon cooling from 1200°C. The combination of high temperatures and reactivity of borosilicate glass with typical containers used in neutron experiments, such as vanadium and niobium, prevented their use here. Thus, methods using sealed thick-walled silica ampoules were developed for the in-situ studies. Unexpectedly, high neutron absorption, low crystal fraction, and high silica container background made quantification difficult for these high temperature measurements. As a follow-up, proof-of-concept measurements were performed on different potential high-temperature container materials, emphasizing crystalline materials so that residual glass in the waste form sample could be more easily analyzed. Room temperature measurements were conducted with a pre-crystallized sample in ‘ideal’ containers stable at low temperatures (i.e., vanadium and thin-wall silica capillaries) and compared to the same measurements in containers stable at high temperatures (i.e, platinum, single crystal sapphire, and thick-walled silica ampoules). Results indicated that Pt is probably the best choice if suitably sealed to prevent contamination from the sample after neutron activation.

36 MATERIALS SCIENCE↗

Compositional Dependence of Solubility/Retention of Molybdenum Oxides in Aluminoborosilicate-Based Model Nuclear Waste Glasses

We present that molybdenum oxides are an integral component of the high-level waste streams being generated from the nuclear reactors in several countries. Although borosilicate glass has been chosen as the baseline waste form by most of the countries to immobilize these waste streams, molybdate oxyanions (MoO 4 2- ) exhibit very low solubility (~1 mol %) in these glass matrices. In the past three to four decades, several studies describing the compositional and structural dependence of molybdate anions in borosilicate and aluminoborosilicate glasses have been reported in the literature, providing a basis for our understanding of fundamental science that governs the solubility and retention of these species in the nuclear waste glasses. However, there are still several open questions that need to be answered to gain an in-depth understanding of the mechanisms that control the solubility and retention of these oxyanions in glassy waste forms. This article is focused on finding answers to two such questions: (1) What are the solubility and retention limits of MoO 3 in aluminoborosilicate glasses as a function of chemical composition? (2) Why is there a considerable increase in the solubility of MoO 3 with incorporation of rare-earth oxides (for example, Nd 2 O 3 ) in aluminoborosilicate glasses? Accordingly, three different series of aluminoborosilicate glasses (compositional complexity being added in a tiered approach) with varying MoO 3 concentrations have been synthesized and characterized for their ability to accommodate molybdate ions in their structure (solubility) and as a glass-ceramic (retention). The contradictory viewpoints (between different research groups) pertaining to the impact of rare-earth cations on the structure of aluminoborosilicate glasses are discussed, and their implications on the solubility of MoO 3 in these glasses are evaluated. Lastly, a novel hypothesis explaining the mechanism governing the solubility of MoO 3 in rare-earth containing aluminoborosilicate glasses has been proposed.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗