THE CBE HARDWARE ACCELERATOR FOR NUMERICAL RELATIVITY: A SIMPLE APPROACH

Laptop Orchestras (LOs) have recently become a very popular mode of musical expression. They engage groups of performers to use ordinary laptop computers as instruments and sound sources in the performance of specially created music software. By using an orchestral metaphor, LOs provide an engaging and challenging environment to experiment with human-computer interaction, network and machine latency, and sound/signal processing. While the LOs at Princeton and Stanford are perhaps the best known, LOs have now been established at many universities in the US and UK, and as private ensembles around the world.Perhaps the biggest challenge for LOs is the distribution, management and control of software across heterogeneous collections of networked computers. Software must be stored and distributed from a central repository, but launched on individual laptops immediately before performance. Each consists of unique combinations of software, user interfaces, and physical devices.Moreover, performers in a Laptop Orchestra can have a complex array of application layers to manage and launch before the start of a specific piece's performance. For example, one work written for our LO requires a bluetooth middleware application which reads gestures from a Wii-mote and converts them into OpenSoundControl [Wright and Freed 1997] messages to be forwarded to a custom Max application, all of which must be launched and configured before any performance. Combine this with the rapid turnaround from one composition to the next during a concert performance, and the problem of preparing members of the laptop orchestra with the appropriate tools for each piece becomes daunting.The GRENDL project leverages proven grid computing frameworks and approaches the Laptop Orchestra as a distributed computing platform for interactive computer music. This allows us to readily distribute software to each laptop in the orchestra depending on the laptop's internal configuration, its role in the composition, and the player assigned to that computer. Using the SAGA framework [Goodale et al.], GRENDL is able to run pre-distribution scripts on a master computer, distribute software to client computers, launch post-distribution scripts on the master computer and launch application scripts on client computers that in turn manage application environments for each composition.SAGA, the Simple API for Grid Applications, is a distributed computing middleware used to distribute, manage and process grid-based applications, typically for scientific research problems in such diverse fields as numerical relativity, computational fluid dynamics and materials science. Its functionality and stability are well regarded within the computational science community and SAGA has become a standard API for grid computing.Our initial experiments have demonstrated that SAGA can be used successfully in a concert environment. The Laptop Orchestra of Louisiana (LOLs) debut concert on April 14, 2010 used a prototype version of GRENDL to manage two of the seven works performed, and GRENDL worked flawlessly.GRENDL proposes to go further than just applying SAGA to the LO environment. We will use tangible and physical objects [Ullmer et al. 2008] to represent individuals, resources, roles and compositions such that GRENDL knows how to distribute software appropriate to the LO's environment, the individuals performing and their role in the composition. By using RFID-embeded objects [Ullmer et al. 2010], master and client computers determine who is at which computers, and what is being performed. Just as a music librarian knows where to place parts for each composition on which music stands, GRENDL will know where to send software and how to launch that software for each composition, laptop and performer.Extending SAGA to work with tangibles and in novel runtime environments, will require extensions to SAGA -- support for new interfaces and instruments -- as well as require some performance engineering in order for commands to be processed with lower latency than traditional distributed systems are designed to tolerate. The trans-disciplinary nature of GRENDL provides potential to shed new light on existing challenges in computational and computer science. The LO setting presents a unique perspective from which to investigate topics such as time-sensitive and dynamic job scheduling, latency-bound interaction, and effective user interfaces for grid computing environments. Some of the first iteration interaction technologies have been developed for distributed computational science applications, and some of what is learned through GRENDL will likely be applicable in that area.

This is a summary of talks and posters presented at GR20 on Numerical Relativity and astrophysical applications. These presentations highlighted a variety of fronts actively being pursued within Numerical Relativity with astrophysics goals.

The current status of convection-permitting regional climate models is reported. First, the outline of regional climate models is described. Next, the representation of clouds in regional climate models is mentioned. Subsequently, the performance of numerical simulations using convection-permitting regional climate models is presented. In particular, the effects of high-resolution topography on the performance of simulations are discussed. Finally, an example of future climate projections using convection-permitting regional climate models is introduced.

In order to improve the accuracy of silicon piezoresistance pressure sensor under different temperatures, temperature compensation should be carried out. A temperature compensation system of pressure sensors is presented. The structure, hardware and software design of the system are introduced. In the hardware design, MAX1452 is used as the main compensation unit. A curve fitting algorithm based on cubic B-spline is adopted in the temperature compensation software design. The accuracy and the production efficiency of the sensors are improved. The system features in low cost, high efficiency and high accuracy. And it has great spreading value

This paper introduces the relevant content of LXI bus technology including its developments, classification, and research status of synchronous triggering of LXI bus. Based on the discussion of hardware triggering techniques, a feasible structure of LXI class A instrument is proposed. Its hardware and software design is described in detail. In hardware design, the M-LVDS drivers are adopted to complete the transmission of triggering events and satisfy the specification of LXI class A instruments. Because of the high performance cast ratio, ARM processor is adopted to be the controller of the instrument. The software scheme is also proposed. The instrument software is designed as a Web server. With a browser, users can access LXI instruments through communicating with IVI-COM drivers. The proposed instrument scheme has the advantages of wider coverage, better succession, longer life cycle, lower cost, greater data throughput and so on.

This paper evaluates the performance of the bioinformatics application ClustalW developed on Cell Broadband Engine(TM) (Cell/B.E.) using a software data cache for SPEs, instead of explicit DMA transfers. The software cache of the SPEs, once it has been configured, provides the capability to access main memory with data-transfer functions that override the need for DMA commands. ClustalW exhibits high spatial locality but little temporal locality. We compare performance of ClustalW with a previous version that uses explicit DMA transfers as a means of communication with the system memory, and provide analysis and results of our comparison.

Due to the absence of symmetries under time and spatial translations in a general curved spacetime, the energy and momentum of matter is not conserved as it is in flat space. This means, for example, that the flux of matter energy through a surface is in general not balanced by an equal increase in the energy of the matter contained within the enclosed volume—there is an additional ‘source’ resulting from the curvature of spacetime acting on the matter (and vice versa). One can calculate this source term and reconcile the flux and energy accumulation over time in an arbitrary volume, although a foliation of the spacetime must be chosen, making the quantities inherently coordinate dependent. Despite this dependence, these quantities are practically useful in numerical relativity simulations for a number of reasons. We provide expressions for general matter sources in a form appropriate for implementation in the Arnowitt Deser Misner decomposition, and discuss several applications in simulations of compact object dynamics and cosmology.

The asymptotic nature of unbound binary-black-hole encounters provides a clean method for comparing different approaches for modeling the two-body problem in general relativity. In this work, we use numerical relativity simulations of black-hole scattering, generated using the Spectral Einstein Code, to explore the self-force and post-Minkowskian expansions of the scattering angle. First, we use a set of unequal-mass simulations to extract the self-force contributions to the scattering angle. Our main result is that using information up to second-order in the symmetric mass ratio (2SF) reproduces numerical relativity within the error bars across the full range of mass-ratios, including equal mass. Next, we compare our numerical relativity results to state-of-the-art post-Minkowskian predictions at larger impact parameters than previously explored. We find good agreement in the weak-field regime and discuss the relative importance of higher order terms.

This paper presents the hardware design of a 16-points 1-D DCT used in the emerging video coding standard HEVC - High Efficiency Video Coding. The 1-D DCT is used by the 16×16 2-D DCT of the HEVC standard. The transforms stage is one of the innovations proposed by HEVC, not only because of the variable size (from 4×4 to 32×32) but also because higher dimension transforms other than the traditional 4×4 and 8×8 are used. The hardware design presented in this work focuses on low cost and high throughput. To achieve such objectives, the 16-points algorithm from HEVC was simplified, so that a more efficient hardware design could be implemented. Some strategies were used during this simplification, such as operations reordering, factoring to compress the length of the operators, multiplications by constant turned into shifts and adds, sub-expressions sharing, among others. The architecture was designed in a fully combinational way in order to reduce hardware overhead. Synthesis results obtained using Altera FPGAs from the Cyclone II and Stratix III families showed hardware resources reduction reaching 72% when compared to an architecture described as a direct transcription of the non-optimized version of the algorithm. Even with a purely combinational implementation, the designed architecture achieved a throughput between 376 Msamples/s and 1.4 Gsamples/s. With these results, the architecture is capable of processing, in the worst case, more than 30 QFHD frames (3840×2160 pixels) per second. Therefore, the architecture is capable of processing videos with significantly high resolutions in real time. To the best of our knowledge, this is the first work in the literature that presents hardware results for the HEVC transforms.

Following from the original Dynamic Adaptive Neural Network Array (DANNA) model, we propose a new digital neuromorphic architecture named DANNA 2. Through this paper, we introduce our new hardware design and software simulator, and we explain how DANNA2 can improve network density, training convergence, and element performance as compared to the DANNA model. We propose two network arrays types catering to FPGA and VLSI use cases. Using simulation results for both control and classification problems, we show improved training convergence and simulation performance.

Purpose This paper aims to study the flow and heat transfer inside a wavy enclosure filled with Cu-water nanofluid under magnetic field effect by parallel implemented meshfree approach. Design/methodology/approach The simulation has been carried out for a two-dimensional model with steady, laminar and incompressible flow of the nanofluid filled inside wavy enclosure in which one of the walls is sinusoidal such that the amplitude (A = 0.15) and number of undulations (n = 2) are fixed. A uniform magnetic field B0 has been applied at an inclination angle γ. The governing equations for the transport phenomena have been solved numerically by implementing element-free Galerkin method (EFGM) with the sequential as well as parallel approach. The effect of various parameters, namely, nanoparticle volume fraction (φ), Rayleigh number (Ra), Hartmann number (Ha) and magnetic field inclination angle (γ) has been studied on the natural convection flow of nanofluid. Findings The results are obtained in terms of average Nusselt number calculated at the cold wavy wall, streamlines and isotherms. It has been observed that the increasing value of Rayleigh number results in increased heat transfer rate while the Hartmann number retards the fluid motion. On the other hand, the magnetic field inclination angle gives rise to the heat transfer rate up to its critical value. Above this value, the heat transfer rate starts to decrease. Originality/value The implementation of the magnetic field and its inclination has provided very interesting results on heat and fluid flow which can be used in the drug delivery where nanofluids are used in many physiological problems. Another important novelty of the paper is that meshfree method (EFGM) has been used here because the domain is irregular. The results have been found to be very satisfactory. In addition, parallelization of the scheme (which has not been implemented earlier in such problems) improves the computational efficiency.

We present techniques for successfully performing numerical relativity simulations of binary black holes with fourth-order accuracy. Our simulations are based on a new coding framework which currently supports higher-order finite differencing for the Baumgarte-Shapiro-Shibata-Nakamura formulation of Einstein's equations, but which is designed to be readily applicable to a broad class of formulations. We apply our techniques to a standard set of numerical relativity test problems, demonstrating the fourth-order accuracy of the solutions. Finally we apply our approach to binary black hole head-on collisions, calculating the waveforms of gravitational radiation generated and demonstrating significant improvements in waveform accuracy over second-order methods with typically achievable numerical resolution.

We present the first surrogate model for gravitational waveforms from the coalescence of precessing binary black holes. We call this surrogate model NRSur4d2s. Our methodology significantly extends recently introduced reduced-order and surrogate modeling techniques, and is capable of directly modeling numerical relativity waveforms without introducing phenomenological assumptions or approximations to general relativity. Motivated by GW150914, LIGO's first detection of gravitational waves from merging black holes, the model is built from a set of $276$ numerical relativity (NR) simulations with mass ratios $q \leq 2$, dimensionless spin magnitudes up to $0.8$, and the restriction that the initial spin of the smaller black hole lies along the axis of orbital angular momentum. It produces waveforms which begin $\sim 30$ gravitational wave cycles before merger and continue through ringdown, and which contain the effects of precession as well as all $\ell \in \{2, 3\}$ spin-weighted spherical-harmonic modes. We perform cross-validation studies to compare the model to NR waveforms \emph{not} used to build the model, and find a better agreement within the parameter range of the model than other, state-of-the-art precessing waveform models, with typical mismatches of $10^{-3}$. We also construct a frequency domain surrogate model (called NRSur4d2s_FDROM) which can be evaluated in $50\, \mathrm{ms}$ and is suitable for performing parameter estimation studies on gravitational wave detections similar to GW150914.

Fine reservoir description is an important prerequisite to ensure the scientific and efficient development of oilfields. Reservoir numerical simulation is an important tool for efficient development. The accuracy of reservoir description determines the reliability of numerical simulation results. After long-term water injection development, the physical properties of the reservoir have undergone great changes. How to identify the time-dependent law of reservoir physical properties and finely characterize them during numerical simulation has been the focus and research difficulty in this field in recent years. This paper combines the results of previous studies on the changes of reservoir physical properties in the later stage of water drive development in Area A with numerical simulation methods. Predecessors realized time-dependent numerical simulation methods include staged numerical simulation and self-developed numerical simulation programs. Although the above two types of methods can achieve the purpose of realizing time-dependent numerical simulation, they all have certain defects. The segmented simulation method requires multiple modeling, heavy workload, multiple manual operations, and poor continuity and non-directional calculation results. Although the self-developed numerical simulation program calculation results can have better continuity and directionality, but this requires a lot of manpower and material resources to carry out the research. The research and development cost is high, the cycle is long, and the simulation accuracy and speed are difficult to reach the level of commercial numerical simulation software in the short term. Based on the above analysis, this paper has developed a simulation program TD-SIM (Time Dependent Simulation) that combines the commercial numerical simulation software ECLIPSE and can realize the time-dependent function of the reservoir permeability field. This program has convenient use and low cost. At the same time, it is necessary to take into account faster simulation speed, higher simulation accuracy, and better continuity and directionality of the simulation results.

Impact of transient flow on bimolecular reactive transport in varying-aperture fractures