Searcharxiv⌕ Search

arXiv subjects

Jun Makino

Publications and source records attributed to Jun Makino.

18 recordsLinked to original sources
arXiv2022

Comments on "Imaging Reanalyses of EHT Data"

On June 14, 2022, the EHT collaboration (hereafter EHTC) made the web page (https://eventhorizontelescope.org/blog/imaging-reanalyses-eht-data) with the title "Imaging Reanalyses of EHT Data," in which they made comments on our recent Miyoshi et al .2022 published in the Astrophysical Journal. We investigated the EHTC comments and found that all of the five points raised by the EHTC are subjective and unsubstantiated claims. Thus they do not prove the correctness of the result of EHTC. Sincerely we hope that the EHTC will publish, not a collection of unsubstantiated claims, but a discussion based on scientific arguments.

astro-ph.HE↗
arXiv2020

Comment on "Individual external dose monitoring of all citizens of Date City by passive dosimeter 5 to 51 months after the Fukushima NPP accident (series): II"

This Letter to the Editor points out several inconsistencies found in the paper written by Makoto Miyazaki and Ryugo Hayano published as J. Radiol. Prot. 37(2017) 623-634. This Letter to the Editor had been in the stage of "is ready to accept" from November 17, 2018 to March 23, 2020 when it has been accepted for publication. Onto this Letter to the Editor, an additional Letter to the Editor on the same paper has been added. The added Letter to the Editor has been provisionally accepted on April 7, 2020.

physics.med-ph↗
arXiv2020

Further comments on "Individual external dose monitoring of all citizens of Date City by passive dosimeter 5 to 51 months after the Fukushima NPP accident (series): 1." : Inconsistencies in Table 1 2014 Q3 and Figure 4f

We point out serious inconsistencies of the first paper of the series, written by Makoto Miyazaki and Ryugo Hayano, which discusses the correlation between the personal doses of the citizens of Date City measured by glass badges with the ambient dose rates measured by six airborne surveys. The last of the six airborne survey was made in the period of 2014 Q3 (from October 2014 to December 2014). The real number of participants of the period is about 14,500; however, in Table 1 2014 Q3 it is written that the number of participants is 21,080 and in Fig. 4f 21,052. We conclude that the analysis of the paper with respect to Table 1 2014 Q3 and Fig. 4f are done without using real correct data and we cannot obtain any meaningful information from the table and figure. Since the period 2014 Q3 is also included in Fig. 5 of the second paper of the series, it is quite possible that Fig. 5 of the second paper is made on the basis of, at least partially, false data and is not reliable.

physics.med-ph↗
arXiv2019

Implementation and Performance of Barnes-Hut N-body algorithm on Extreme-scale Heterogeneous Many-core Architectures

In this paper, we report the implementation and measured performance of our extreme-scale global simulation code on Sunway TaihuLight and two PEZY-SC2 systems: Shoubu System B and Gyoukou. The numerical algorithm is the parallel Barnes-Hut tree algorithm, which has been used in many large-scale astrophysical particle-based simulations. Our implementation is based on our FDPS framework. However, the extremely large numbers of cores of the systems used (10M on TaihuLight and 16M on Gyoukou) and their relatively poor memory and network bandwidth pose new challenges. We describe the new algorithms introduced to achieve high efficiency on machines with low memory bandwidth. The measured performance is 47.9, 10.6 PF, and 1.01PF on TaihuLight, Gyoukou and Shoubu System B (efficiency 40\%, 23.5\% and 35.5\%). The current code is developed for the simulation of planetary rings, but most of the new algorithms are useful for other simulations, and are now available in the FDPS framework.

astro-ph.IM↗
arXiv2016

Implementation and evaluation of data-compression algorithms for irregular-grid iterative methods on the PEZY-SC processor

Iterative methods on irregular grids have been used widely in all areas of comptational science and engineering for solving partial differential equations with complex geometry. They provide the flexibility to express complex shapes with relatively low computational cost. However, the direction of the evolution of high-performance processors in the last two decades have caused serious degradation of the computational efficiency of iterative methods on irregular grids, because of relatively low memory bandwidth. Data compression can in principle reduce the necessary memory memory bandwidth of iterative methods and thus improve the efficiency. We have implemented several data compression algorithms on the PEZY-SC processor, using the matrix generated for the HPCG benchmark as an example. For the SpMV (Sparse Matrix-Vector multiplication) part of the HPCG benchmark, the best implementation without data compression achieved 11.6Gflops/chip, close to the theoretical limit due to the memory bandwidth. Our implementation with data compression has achieved 32.4Gflops. This is of course rather extreme case, since the grid used in HPCG is geometrically regular and thus its compression efficiency is very high. However, in real applications, it is in many cases possible to make a large part of the grid to have regular geometry, in particular when the resolution is high. Note that we do not need to change the structure of the program, except for the addition of the data compression/decompression subroutines. Thus, we believe the data compression will be very useful way to improve the performance of many applications which rely on the use of irregular grids.

cs.MS↗
arXiv2007

Distributed N-body Simulation on the Grid Using Dedicated Hardware

We present performance measurements of direct gravitational N -body simulation on the grid, with and without specialized (GRAPE-6) hardware. Our inter-continental virtual organization consists of three sites, one in Tokyo, one in Philadelphia and one in Amsterdam. We run simulations with up to 196608 particles for a variety of topologies. In many cases, high performance simulations over the entire planet are dominated by network bandwidth rather than latency. With this global grid of GRAPEs our calculation time remains dominated by communication over the entire range of N, which was limited due to the use of three sites. Increasing the number of particles will result in a more efficient execution. Based on these timings we construct and calibrate a model to predict the performance of our simulation on any grid infrastructure with or without GRAPE. We apply this model to predict the simulation performance on the Netherlands DAS-3 wide area computer. Equipping the DAS-3 with GRAPE-6Af hardware would achieve break-even between calculation and communication at a few million particles, resulting in a compute time of just over ten hours for 1 N -body time unit. Key words: high-performance computing, grid, N-body simulation, performance modelling

astro-ph↗
arXiv2007

The core radius of a star cluster containing a massive black hole

(Simplified) We present a theoretical framework which establishes how the core radius of a star cluster varies with the mass of an assumed central black hole. Our result is that the ratio of core to half-mass radius varies as the 3/4 power of the ratio of the black hole to cluster mass, when the system is well relaxed. The theory compares favourably with a number of simulations of this problem, which extend to black hole masses of order 10% of the cluster mass. Though strictly limited as yet to clusters with stars of equal mass, our conclusion strengthens the view that clusters with large core radii are the most promising candidates in which to find a massive black hole.

astro-ph↗
arXiv2006

Star Cluster Ecology: VII The evolution of young dense star clusters containing primordial binaries

We study the first 100Myr of the evolution of isolated star clusters initially containing 144179 stars, including 13107 (10%) primordial hard binaries. Our calculations include the effects of both stellar and binary evolution. Gravitational interactions among the stars are computed by direct N-body integration using high precision GRAPE-6 hardware. The evolution of the core radii and central concentrations of our simulated clusters are compared with the observed sample of young (about 100Myr) star clusters in the large Magellanic cloud. Even though our simulations start with a rich population of primordial binaries, core collapse during the early phase of the cluster evolution is not prevented. Throughout the simulations, the fraction of binaries remains roughly constant (about 10%). Due to the effects of mass segregation the mass function of intermediate-mass main-sequence stars becomes as flat as $α=-1.8$ in the central part of the cluster (where the initial Salpeter mass function had $α=-2.35$). About 6--12% of the neutron stars were retained in our simulations; the fraction of retained black holes is 40--70%. In each simulation about three neutron stars become members of close binaries with a main-sequence companion. Such a binary will eventually become an x-ray binary, when the main-sequence star starts to fill its Roche lobe. Black holes are found more frequently in binaries; in each simulated cluster we find about 11 potential x-ray binaries containing a black hole. Abstract abbreviated....

astro-ph↗
arXiv2005

A New Time-Symmetric Block Time-Step Algorithm for N-Body Simulations

Time-symmetric integration schemes share with symplectic schemes the property that their energy errors show a much better behavior than is the case for generic integration schemes. Allowing adaptive time steps typically leads to a loss of symplecticity. In contrast, time symmetry can be easily maintained, at least for a continuous choice of time step size. In large-scale N-body simulations, however, one often uses block time steps, where all time steps are forced to take on values as powers of two. This greatly facilitates parallelization, and hence code efficiency. Straightforward implementation of time-symmetry, translated to block time steps, faces significant hurdles. For example, iteration can lead to oscillatory behavior, and even when such behavior is suppressed, energy errors show a linear drift in time. We present an approach that circumvents these problems.

astro-ph↗
arXiv2004

Dynamics of Intermediate Mass Black Holes in Star Clusters

We have followed the evolution of multi-mass star clusters containing massive central black holes by N-body simulations on the GRAPE6 computers of Tokyo University. We find a strong cluster expansion and significant structural changes of the clusters. Star clusters with IMBHs have power-law density profiles $rho ~ r^{-alpha}$ with slopes $alpha=1.55$ inside the influence sphere of the central black hole. This leads to a constant density profile of bright stars in projection, which rules out the presence of intermediate mass black holes in core collapse clusters. If the star clusters are surrounded by a tidal field, a central IMBH speeds up the destruction of the cluster until a remnant of a few hundred stars remains, which stays bound to the IMBH for a long time. We also discuss the efficiency of different detection mechanisms for finding IMBHs in star clusters.

astro-ph↗
arXiv2003

Star cluster ecology V: Dissection of an open star cluster--spectroscopy

We have modeled in detail the evolution of rich open star clusters such as the Pleiades, Praesepe and Hyades, using simulations that include stellar dynamics as well as the effects of stellar evolution. The dynamics is modeled via direct N-body integration, while the evolution of single stars and binaries is followed through the use of fitting formulae and recipes. The feedback of stellar and binary evolution on the dynamical evolution of the stellar system is taken into account self-consistently. Our model clusters dissolve in the tidal field of the Galaxy in a time span on the order of a billion years. The rate of mass loss is rather constant, about 1Msun per million years. The binary fraction at first is nearly constant in time, then increases slowly near the end of a cluster's lifetime. For clusters which are more than about 100 million years old the fractions of stars in the form of binaries, giants and collision products in the inner few core radii are considerably higher than in the outer regions, beyond the cluster's half mass radius. When stars with masses >2Msun escape from the cluster, they tend to do so with velocities higher than average....

astro-ph↗
arXiv2001

Theory in a Virtual Observatory

During the last couple of years, observers have started to make plans for a Virtual Observatory, as a federation of existing data bases, connected through levels of software that enable rapid searches, correlations, and various forms of data mining. We propose to extend the notion of a Virtual Observatory by adding archives of simulations, together with interactive query and visualization capabilities, as well as ways to simulate observations of simulations in order to compare them with observations. For this purpose, we have already organized two small workshops, earlier in 2001, in Tucson and Aspen. We have also provided concrete examples of theory data, designed to be federated with a Virtual Observatory. These data stem from a project to construct an archive for our large-scale simulations using the GRAPE-6 (a 32-Teraflops special purpose computer for stellar dynamics). We are constructing interfaces by which remote observers can observe these simulations. In addition, these data will enable detailed comparisons between different simulations.

astro-ph↗
arXiv2001

Immersive 4D Interactive Visualization of Large-Scale Simulations

In dense clusters a bewildering variety of interactions between stars can be observed, ranging from simple encounters to collisions and other mass-transfer encounters. With faster and special-purpose computers like GRAPE, the amount of data per simulation is now exceeding 1TB. Visualization of such data has now become a complex 4D data-mining problem, combining space and time, and finding interesting events in these large datasets. We have recently starting using the virtual reality simulator, installed in the Hayden Planetarium in the American Museum for Natural History, to tackle some of these problem. This work (http://www.astro.umd.edu/nemo/amnh/) reports on our first ``observations'', modifications needed for our specific experiments, and perhaps field ideas for other fields in science which can benefit from such immersion. We also discuss how our normal analysis programs can be interfaced with this kind of visualization.

astro-ph↗
arXiv1998

High-Performance Special-Purpose Computers in Science

The next decade will be an exciting time for computational physicists. After 50 years of being forced to use standardized commercial equipment, it will finally become relatively straightforward to adapt one's computing tools to one's own needs. The breakthrough that opens this new era is the now wide-spread availability of programmable chips that allow virtually every computational scientist to design his or her own special-purpose computer.

astro-ph↗
arXiv1998

Astrophysics on the GRAPE Family of Special Purpose Computers

The GRAPE-4, the world's fastest computer in 1995-1997, has produced some major scientific results, through a wide diversity of large-scale simulations in astrophysics. Applications have ranged from planetary formation, through the evolution of star clusters and galactic nuclei, to the formation of galaxies and clusters of galaxies.

astro-ph↗
arXiv1997

Smooth Particle Hydrodynamics: Models, Applications, and Enabling Technologies

We present the results from a two-day study in which we discussed various implementations of Smooth Particle Hydrodynamics (SPH), one of the leading methods used across a variety of areas of large-scale astrophysical simulations. In particular, we evaluated the suitability of designing special hardware extensions, to further boost the performance of the high-end general purpose computers currently used for those simulations. We considered a range of hybrid architectures, consisting of a mix of custom LSI and reconfigurable logic, combining the extremely high throughput of Special-Purpose Devices (SPDs) with the flexibility of reconfigurable structures, based on Field Programmable Gate Arrays (FPGAs). The main findings of our workshop consist of a clarification of the decomposition of the computational requirements, together with specific estimates for cost/performance improvements that can be obtained at each stage in this decomposition, by using enabling hardware technology to accelerate the performance of general purpose computers.

astro-ph↗