摘要 :
Irregular access patterns are a major problem for today's optimizing compilers. In this paper, a novel approach will be presented that enables transformations that were designed for regular loop structures to be applied to linked ...
展开
Irregular access patterns are a major problem for today's optimizing compilers. In this paper, a novel approach will be presented that enables transformations that were designed for regular loop structures to be applied to linked list data structures. This is achieved by linearizing access to a linked list, after which further data restructuring can be performed. Two subsequent optimization paths will be considered: annihilation and sublimation, which are driven by the occurring regular and irregular access patterns in the applications. These intermediate codes are amenable to traditional compiler optimizations targeting regular loops. In the case of sublimation, a run-time step is involved which takes the access pattern into account and thus generates a data instance specific optimized code. Both approaches are applied to a sparse matrix multiplication algorithm and an iterative solver: preconditioned conjugate gradient. The resulting transformed code is evaluated using the major compilers for the x86 platform, GCC and the Intel C compiler.
收起
摘要 :
We explored correlation in diet trends for five piscivorous predators that reproduce on the Pribilof Islands as illustrative of the shifting structure of the Bering Sea ecosystem. We evaluated the size and species of prey consumed...
展开
We explored correlation in diet trends for five piscivorous predators that reproduce on the Pribilof Islands as illustrative of the shifting structure of the Bering Sea ecosystem. We evaluated the size and species of prey consumed by adult female and juvenile northern fur seals (Callorhinus ursinus) and adults and chicks of black-legged kittiwakes (Rissa tridactyla), red-legged kittiwakes (Rissa brevirostris), thick-billed murres (Uria lomvia), and common murres (Una aalge) from data collected between July and October 1960-2000. Sample sources included stomachs from seals and seabirds collected on pelagic foraging grounds in the eastern Bering Sea, seal scats from rookeries and seabird regurgitations and whole prey from nest sites on St. Paul and St. George Islands of the Pribilof Island archipelago. Typical prey included small fish and invertebrates (≤ 20 cm for seals and ≤ 12 cm for seabirds) that concentrate along frontal boundaries of the continental shelf/slope and in the epi-pelagic zone. Squids and fishes including walleye pollock (Theragra chalcogramma), capelin (Mallotus villosus), and sand lance (Ammodytes hexapterus) were variably important in the diet of all five predators. Some prey, such as capelin, were principal in predator diets during the 1960s (seals) and into the early 1980s (seabirds), but declined or disappeared from all predator diets thereafter while others, such as walleye pollock, occurred with increasing frequency from the 1970s forward. As the number of individuals consuming walleye pollock increased, the overall volume of pollock in seabird diets declined. This decline was coincident with a decrease in the age and body size of pollock consumed by both seabirds and fur seals. Squid and pollock were negatively correlated in the diets of their primary consumers, northern fur seals (Pearson's coefficient -0.71, p = 0.016) and thick-billed murres (Pearson's coefficient = -0.74, p = 0.015) from the 1970s forward. Inter-island variation in diet was evident to varying degrees for all predators, with a prevalence offish on St. Paul Island and invertebrates on St. George Island. Bayesian time-series analysis of synthesized data described significant temporal cross-correlation in diet among northern fur seals, red- and black-legged kittiwakes, and thick-billed murres. For all correlated predators except common murres, beta-binomial modeling indicated that trends in the occurrence of four of the five primary prey (sand lance, capelin, squid, and pollock) evaluated, were significantly associated with eastern Bering Sea time-series trends in sea surface temperature, ice retreat or a combination of both. Data synthesis highlighted potential competition and a scenario for the effects of an altered prey field on the population stability of predators. The association between correlated diet changes among predators and indices of oceanographic shifts in the 1970s and the 1990s allow scrutiny of hypotheses concerning causal mechanisms in population declines.
收起
摘要 :
The motion of a vortex domain wall in a ferromagnetic strip of submicron width under the influence of an external magnetic field exhibits three distinct dynamical regimes. In a viscous regime at low fields the wall moves rigidly w...
展开
The motion of a vortex domain wall in a ferromagnetic strip of submicron width under the influence of an external magnetic field exhibits three distinct dynamical regimes. In a viscous regime at low fields the wall moves rigidly with a velocity proportional to the field. Above a critical field the viscous motion breaks down, giving way to oscillations accompanied by a slow drift of the wall. At still higher fields the drift velocity starts rising with the field again but with a much lower mobility dv/dH than in the viscous regime. To describe the dynamics of the wall, we use the method of collective coordinates that focuses on soft modes of the system. By retaining two soft modes, parametrized by the coordinates of the vortex core, we obtain a simple description of the wall dynamics at low and intermediate applied fields that applies to both the viscous and oscillatory regimes below and above the breakdown. The calculated dynamics agrees well with micromagnetic simulations at low and intermediate values of the driving field. In higher fields, additional modes become soft and the two-mode approximation is no longer sufficient. We explain some of the significant features of vortex-domain-wall motion in high fields through the inclusion of additional modes associated with the half antivortices on the strip edge.
收起
摘要 :
Time-resolved optical spectroscopy has been used to measure the shock pressure steadiness, emissivity, and temperature of liquid deuterium shocked to 22-90 GPa. The shock was produced using magnetically accelerated flyer plate imp...
展开
Time-resolved optical spectroscopy has been used to measure the shock pressure steadiness, emissivity, and temperature of liquid deuterium shocked to 22-90 GPa. The shock was produced using magnetically accelerated flyer plate impact, and spectra were acquired with a suite of four fiber-optic-coupled spectrometers with streak camera detectors. The shock pressure changes by an average of -1.2% over the 10-30 ns cell transit time, determined from the relative changes in the shock front self-emission with time. The shock front reflectivity was measured from 5140 A and 5320 A laser light reflected from the D_2 shock. The emissivity inferred from the reflectivity measurements was in reasonably good agreement with quantum molecular dynamics simulation predictions. The spectral radiance wavelength dependence was found to agree well (average normalized χ~2=1.6) with a Planckian multiplied by the emissivity. The shock front temperature was determined from the emissivity and the wavelength-dependent shock self-emission. Thirty-seven temperature measurements spanning the 22-90 GPa range were accumulated. The large number of temperature measurements enables a comparison of the scatter in the data with expectations for a Gaussian distribution. This facilitates determination of uncertainties that incorporate both apparatus contributions and otherwise unquantified systematic effects that cause self-emission variations from one experiment to another. Agreement between temperatures determined from the absolute spectral radiance and from the relative shape of the spectrum further substantiates the absence of systematic biases. The weighted mean temperature uncertainties were as low as ±3-4%, enabling the discrimination between competing models for the D_2 equation of state (EOS). The temperature results agree well with models that predict a maximum compression of ～4.4. Softer models that predict approximately sixfold compression are inconsistent with the data to a very high statistical confidence level. Previous analysis [D. Saumon and T. Guillot, Astrophys. J. 609, 1170 (2004)] of Jupiter's internal structure has shown that the core mass is restricted to be less than approximately three times the mass of the Earth, if EOS models consistent with these temperature measurements are employed.
收起
摘要 :
We review how a magnetic-resonance force microscope (MRFM) can be applied to perform ferromagnetic resonance spectroscopy of individual submicron-size samples. We restrict our attention to a thorough study of the spin-wave eigenmo...
展开
We review how a magnetic-resonance force microscope (MRFM) can be applied to perform ferromagnetic resonance spectroscopy of individual submicron-size samples. We restrict our attention to a thorough study of the spin-wave eigenmodes excited in Permalloy (Py) disks patterned out of the same 43.3-nm-thin film. The disks have a diameter of either 1.0 or 0.5 μm and are quasisaturated by a perpendicularly applied magnetic field. It is shown that quantitative spectroscopic information can be extracted from the MRFM measurements. In particular, the data are extensively compared with complementary approximate models of the dynamical susceptibility: (ⅰ) a two-dimensional analytical model, which assumes a homogeneous magnetization dynamics along the thickness, and ⅱ) a full three-dimensional micromagnetic simulation, which assumes a homogeneous magnetization dynamics below a characteristic length scale c and approximates the cylindrical sample volume by a discretized representation with regular cubic mesh of lateral size c=3.9 nm. In our analysis, the distortions due to a breaking of the axial symmetry are taken into account; both models incorporating the possibility of a small misalignment between the applied field and the normal of the disks.
收起
摘要 :
The reflood rate in a core is an important parameter for core cooling during a large-break loss-of-coolant-accident (LBLOCA) reflood period, and it strongly depends on the thermal-hydraulic conditions in the down-comer. During thi...
展开
The reflood rate in a core is an important parameter for core cooling during a large-break loss-of-coolant-accident (LBLOCA) reflood period, and it strongly depends on the thermal-hydraulic conditions in the down-comer. During this period, downcomer boiling has an important influence on the transient behavior of a postulated LBLOCA because it can degrade the hydraulic head in a downcomer and consequently affect the reflood flow rate for core cooling. Although it is recognized that downcomer boiling is critical to correctly predict the reflood phenomena of an LBLOCA transient, especially for a direct vessel injection adapted system like the advanced power reactor APR1400, the amount of experimental data and code assessment in this area is relatively limited. To improve the state of knowledge relative to downcomer boiling, a test program at the Downcomer Boiling (DOBO) facility is progressing for the reflood phase of a postulated LBLOCA. The DOBO facility was designed to meet a full scale for the height and gap of a reactor downcomer. The DOBO test revealed a strong multidimensional boiling behavior, which induces the need for performance evaluation of the best-estimate codes that are used to analyze a nuclear reactor's thermal-hydraulic safety, since they have mostly been used for one-dimensional system behavior. In this study, RELAP, MARS, and TRACE are evaluated by using measured two-phase-flow data. Based on the assessments, the modeling capability and weak points of the safety analysis codes are addressed for multidimensional downcomer boiling phenomena. Two models for a downcomer are considered to assess the codes for the DOBO tests, which are also applied to a plant analysis.
收起
摘要 :
An exact hypergeometric implicit solution of the 1D steady-state heat conduction problem for a straight fin of constant cross-section is used to calculate the dependence of the main dimensionless fin parameters, specifically, base...
展开
An exact hypergeometric implicit solution of the 1D steady-state heat conduction problem for a straight fin of constant cross-section is used to calculate the dependence of the main dimensionless fin parameters, specifically, base thermal conductance G and thermo-geo-metrical fin parameter N on T_e, the ratio of the fin tip to fin base temperature excesses. The straight plate fin (SPF) and cylindrical pin fin (CPF) with an insulated tip (INT) and non-insulated tip (NINT) are optimized. The local heat transfer coefficient (HTC) is assumed to vary as power function of the local fin excess temperature with arbitrary value of exponent n in the range of -0.5 ≤ n ≤ 5. Every curve from G vs T_e set at given n for a fin with an INT is shown to have a single global maximum G = G_(opt)~* at T_e = T_(e,opt)~* and corresponding N = N_(opt)~*, i.e. the main optimum parameters depend only on exponent n. Every curve from G vs T_e set for a fin with a NINT depends, in addition, on the complex fin tip parameter B_ω. These curves have the local maximum and minimum points. As B_ω increases these points approach each other and at B_ω = B~(**) merge. The corresponding curve G vs T_e has the only inflection point. The main optimum parameters of a fin with an INT and inflection point parameters of a fin with a NINT are approximated by general homographic function of n. Each main optimum parameter of a fin with a NINT is expressed as a product of the corresponding parameter of this fin with an INT and a correction factor approximated by the generalized closed-form formula. The results of the study are presented in form of dimensionless explicit relations, tables and plots which are well suited for the thermal design of optimum fins.
收起
摘要 :
A closed-form inverse solution of the 1-D heat conduction problem for a single fin or spine of constant cross section with an insulated tip is generalized to account for the effect of the tip heat loss. The heat transfer coefficie...
展开
A closed-form inverse solution of the 1-D heat conduction problem for a single fin or spine of constant cross section with an insulated tip is generalized to account for the effect of the tip heat loss. The heat transfer coefficient (HTC) is assumed to exhibit the power-law type dependence on the local excess temperature with arbitrary value of the exponent n in the range of -0.5 ≤ n ≤ 5. The form of the obtained inverse solution is the same as the one for a fin with an insulated tip. However, in addition to the dimensionless fin tip temperature T_e and n, the fin parameter N also depends on the complex parameter ω~2Bi. Using the inversion of this solution and a linearization procedure, the recurrent direct solution with a high convergence rate is derived. Based on the latter, the explicit direct closed-form solution for the accurate determination of the temperature distribution along a fin height at the given values of N, n, and ω~2Bi is obtained. This allows one to determine the base thermal conductance G of the straight plate fin (SPF) and cylindrical pin fin (CPF). The relations between the fin parameters are systematized and collected in two tables for the SPF and CPF. They permit one to determine the arbitrary dimension-less geometrical or thermal fin parameter at given value of any other of its parameters and prescribed or calculated values of the main fin parameter(s) N or (and) G.
收起
摘要 :
110 and 25/200/120 showed no significant difference in TDF contents compared with its raw rice. Multiple regression models showed that feed moisture had the most pronounced effect on extrudate qualities, followed by barrel tempera...
展开
110 and 25/200/120 showed no significant difference in TDF contents compared with its raw rice. Multiple regression models showed that feed moisture had the most pronounced effect on extrudate qualities, followed by barrel temperature and screw speed.
收起
摘要 :
Groundnut burfi with or without sorbic acid (0.3%) based on roasted groundnut, sugar, milk powder, condensed milk and flavoring materials was developed. The changes in quality of groundnut burfi packed in polypropylene (PP, 75 μ)...
展开
Groundnut burfi with or without sorbic acid (0.3%) based on roasted groundnut, sugar, milk powder, condensed milk and flavoring materials was developed. The changes in quality of groundnut burfi packed in polypropylene (PP, 75 μ) and metallized polyester (12 μ) low density/linear low density (MP, 75 μ) were monitored during storage in order to assess the shelf life. The sample without sorbic acid spoiled within 30 days of storage due to mold growth and fermented odor. Groundnut burfi containing sorbic acid did not support any microbial growth during storage of up to 8 months. Groundnut burfi remained stable and acceptable up to 6 months and 8 months under ambient temperatures (15-34C) in PP and MP pouches, respectively. Peroxide and thiobarbituric acid values were higher for the product packed in PP than for the one packed in MP pouches. Oleic acid was the major fatty acid present in fat extracted from groundnut burfi followed by linoleic and palmitic acids. Sorbic acid degraded during storage of groundnut burfi and the rate of degradation was higher for samples packed in PP than those packed in MP pouches.
收起
