Shibata N.,Shiga University
Computer Science - Research and Development | Year: 2010
Data-parallel architectures like SIMD (Single Instruction Multiple Data) or SIMT (Single Instruction Multiple Thread) have been adopted in many recent CPU and GPU architectures. Although some SIMD and SIMT instruction sets include double-precision arithmetic and bitwise operations, there are no instructions dedicated to evaluating elementary functions like trigonometric functions in double precision. Thus, these functions have to be evaluated one by one using an FPU or using a software library. However, traditional algorithms for evaluating these elementary functions involve heavy use of conditional branches and/or table look-ups, which are not suitable for SIMD computation. In this paper, efficient methods are proposed for evaluating the sine, cosine, arc tangent, exponential and logarithmic functions in double precision without table look-ups, scattering from, or gathering into SIMD registers, or conditional branches. We implemented these methods using the Intel SSE2 instruction set to evaluate their accuracy and speed. The results showed that the average error was less than 0.67 ulp, and the maximum error was 6 ulps. The computation speed was faster than the FPUs on Intel Core 2 and Core i7 processors. © 2010 Springer-Verlag.
Matsushita K.,Shiga University |
Yamane F.,Kobe University
Energy Economics | Year: 2012
Under the scheme of the Kyoto Protocol, there are plans for the efficient reduction of carbon dioxide emissions. In the electric power sector, nuclear power generation, which emits no carbon dioxide in the process of generating electricity, has come under scrutiny. However, this energy produces a new environmental issue: the disposal of radioactive waste. First, we derive shadow prices of carbon dioxide and low-level waste as marginal abatement costs in the case of the electric power sector in Japan, employing a directional output distance function. It is found that the shadow prices are US$39 per tonne for carbon dioxide and US$1531 per liter for low-level waste. Secondly, we calculate the indirect Morishima elasticity between carbon dioxide and low-level waste in order to identify their substitutability, and it is found that the substitution of low-level waste for carbon dioxide is easier than the reverse. This result suggests that, with the amount of generated electricity fixed, carbon dioxide can be substituted more easily by low-level waste when the relative price of carbon dioxide increases, for example, as a result of implementation of a carbon dioxide tax or an emissions trading system. © 2011 Elsevier B.V.
Kayama M.,Okayama University of Science |
Nakano S.,Shiga University |
Nishido H.,Okayama University of Science
American Mineralogist | Year: 2010
Alkali feldspars in syenite from the Cerro Balmaceda pluton in the Patagonian Andes, Chile, show various petrographic microtextures formed during the magmatic to high- and low-temperature hydrothermal stages in which cathodoluminescence (CL) shows a wide range of blue, violet, and pink to red colors with variable brightness. Their CL spectra exhibit two emission bands: one at 405-420 nm in the blue region and the other at 700-760 nm in the red-infrared (IR) region. Asymmetrically shaped spectral peaks in energy units suggest overlapping of each individual emission, which corresponds to various luminescence centers. Blue emission bands were separated into two spectral peaks fitted by Gaussian curves centered at 3.055-3.076 and 2.815-2.845 eV. A positive correlation is found between emission intensities at 3.055-3.076 eV and TiO 2 contents, suggesting the activation of a Ti 4+ impurity as an emission center. The intensities at 2.815-2.845 eV, where clear and featureless feldspar (CF; not affected by hydrothermal metasomatism) is shown under optical microscopy, which have intensities appreciably higher than those showing patched microperthite (PMP), formed during low-temperature hydrothermal reactions, correlate reciprocally with the intensities of red-IR emission caused by a Fe 3+ impurity center. The peak at 2.815-2.845 eV can be attributed to oxygen defects associated with Al-O-Al and Al-O-Ti bridges. Most of the areas show CL emissions at 700-760 nm in the red-IR region, in which intensities increase with an increase in Fe 2O 3 contents as impurities. The Fe 3+ ion acts as an activator for the red-IR emission. The Ab-rich and Or-rich phases of PMP have emission components at 1.644 eV (754 nm) and 1.727 eV (717 nm), respectively. The red-IR emission from CF consists of emission components at 1.677 eV (739 nm) and 1.557 eV (796 nm), according to an Fe 3+ impurity center in the Or-rich phase and in the Ab-rich phase as cryptoperthite, respectively. Both components are centered at a wavelength longer than the emission band of Ab-rich and Or-rich phases of PMP, suggesting a change in configurational state around the Fe3+ ion from the T2 to the T1 site by low-temperature hydrothermal metasomatic reactions. Accordingly, the peak positions of the red-IR emission are controlled by the ordering state of Fe 3+ ion into the T1 site, the existence of multiphase perthite and chemical composition.
Hattori A.,Shiga University
Behavioral Ecology | Year: 2012
I examined why anemonefishes, Amphiprion ocellaris, Amphiprion frenatus, and Amphiprion perideraion, which inhabit single host anemones, are monogamous and protandrous. Because they live in small groups (≤6 individuals) with a size hierarchy, they might have the environmental potential for polygyny. If the dominant fish were male, he could monopolize mating opportunities, and subordinates should be female. Female fecundity increases with body size, so that group body size composition largely influences his reproductive success. First, I developed an optimization model to predict the body size composition based on the carrying capacity of the host (C) and a fixed body size ratio (γ) or difference (δ) between individuals adjacent in rank. The γ and δ were assumed to be necessary for subordinates to avoid fatal eviction from the group. The model using δ and C could predict the body size composition of the 3 species. Next, I incorporated nonlinear female fecundity functions into the model to evaluate whether polygyny or monogamy was the better mating system for the dominant fish. I examined relationships between gonad weight and body size of A. frenatus and A. perideraion to estimate the functions. Assuming large δs and the nonlinear functions, dominant fish could have higher reproductive success in monogamous mating systems. The model also indicated that where C is limited, a larger δ resulted in one large, one small, and several very small subordinate individuals. This combination of body size composition model and nonlinear female fecundity functions can explain the function of protandry and monogamy under limited shelter space. © 2012 The Author.
Mori K.,Shiga University |
Perrings C.,Arizona State University
Science of the Total Environment | Year: 2012
This paper presents a model of the problem on floodplain development, exploring the conditions that are both necessary and sufficient for development to be optimal. The model is calibrated for a particular catchment, the Ouse catchment in the United Kingdom, and is used both to estimate the expected impact of floodplain development and to explore the impact of alternative policy instruments. We find that the use of price-based instruments that signal the expected flood damage cost of floodplain development has the potential to lead to outcomes close to the social optimum. The finding is robust to two types of uncertainty: model error about the relation between precipitation and flood-risk and measurement error about the benefits of developed floodplains. © 2012 Elsevier B.V..