scholarly journals FPGA Fast Simulation of Tsunami Wave Propagation

2021 ◽  
pp. 14-25
Author(s):  
М. М. Лаврентьев ◽  
К. Ф. Лысаков ◽  
А. Г. Марчук ◽  
К. К. Облаухов
Keyword(s):  
Wave Propagation ◽  
Shallow Water ◽  
Tsunami Wave ◽  
Quantitative Estimation ◽  
Kamchatka Peninsula ◽  
Modern Computer ◽  
The Pacific ◽  
Field Programmable ◽  
Significant Performance ◽  
Programmable Gate Arrays

В данной статье рассматривается решение задачи быстрой численной оценки высоты волн цунами от гипотетического очага вдоль тихоокеанского побережья полуострова Камчатка и Курильских островов. Мы фокусируемся на очень быстром (практически в режиме поступления данных) численном моделировании распространения волны цунами на основе ПК в соответствии с классическим приближением теории мелкой воды. Существенный прирост производительности достигается за счет использования преимуществ современных компьютерных архитектур, а именно вентильных матриц, программируемых пользователем (Field Programmable Gate Array – FPGA). Разностная схема Мак-Кормака второго порядка аппроксимации для решения системы дифференциальных уравнений мелкой воды [1] реализована на чипе FPGA в составе платы, специально разработанной авторами для решения этой задачи [2, 3]. Численные тесты показывают, что для расчета 3600 шагов по времени распространения волны цунами в расчетной области размером приблизительно 2000х2000 км (3120х2400 расчетных узлов) требуется всего несколько секунд для моделирования цунами от модельного источника волны цунами на сетке с пространственным шагом около 900 м. Созданный на базе FPGA спецвычислитель был также протестирован по точности сравнением с аналитическими решениями, полученными Ан. Марчуком [4, 5] для некоторых модельных топографий дна. The study offers a fast quantitative estimation of tsunami wave heights coming from a hypothetical source along the Pacific coast of the Kamchatka Peninsula and the Kuril Islands. We focus on a very fast (virtually real-time) PC simulation of tsunami wave propagation using the classical approximation of the shallow water theory. Significant performance gains are achieved by taking advantage of modern computer architectures, namely Field Programmable Gate Arrays (FPGAs). The McCormack difference scheme of the second order of approximation for solving the system of shallow water differential equations [1] is implemented with an FPGA chip on a custom PCB designed by the authors [2, 3]. Numerical tests indicate that it takes only a few seconds to simulate a tsunami wave from a simulated source on a 900 m spacing grid to analyze 3,600 time increments of propagation of the tsunami wave propagation in about 2000x2000 km area (3,120x2,400 nodes.) The customized FPGA computer was also tested for accuracy by comparing with the analytical solutions obtained by Marchuk [4, 5] for some reference bottom topographies.  

scholarly journals Hardware Acceleration of Tsunami Wave Propagation Modeling in the Southern Part of Japan

2020 ◽  
Vol 10 (12) ◽  
pp. 4159
Author(s):  
Mikhail Lavrentiev ◽  
Konstantin Lysakov ◽  
Andrey Marchuk ◽  
Konstantin Oblaukhov ◽  
Mikhail Shadrin
Keyword(s):  
Wave Propagation ◽  
Tsunami Wave ◽  
Water System ◽  
Hardware Acceleration ◽  
Tsunami Source ◽  
Propagation Modeling ◽  
Field Programmable ◽  
Shallow Water System ◽  
Industrial Unit ◽  
Wave Propagation Modeling

In order to speed up the calculation of tsunami wave propagation, the field-programmable gate array (FPGA) microchip is used. This makes it possible to achieve valuable performance gain with a modern regular personal computer. The two half-step MacCormack scheme was used herein for numerical approximation of the shallow water system. We studied the distribution of tsunami wave maximal heights along the coast of the southern part of Japan. In particular, the dependence of wave maximal heights on the particular tsunami source location was investigated. Synthetic 100 × 200 km sources have realistic parameters corresponding to this region. As observed numerically, only selected parts of the entire coast line are subject to dangerous tsunami wave amplitudes. The particular locations of such areas strongly depend on the location of the tsunami source. However, the extreme tsunami heights in some of those areas can be attributed to local bathymetry. The proposed hardware acceleration to compute tsunami wave propagation can be used for rapid (say, in a few minutes) tsunami wave danger evaluation for a particular village or industrial unit on the coast.

scholarly journals Design of Flexible Hardware Accelerators for Image Convolutions and Transposed Convolutions

2021 ◽  
Vol 7 (10) ◽  
pp. 210
Author(s):  
Cristian Sestito ◽  
Fanny Spagnolo ◽  
Stefania Perri
Keyword(s):  
Computer Vision ◽  
Super Resolution ◽  
Feature Maps ◽  
Operating Modes ◽  
Gate Arrays ◽  
Modern Computer ◽  
Field Programmable ◽  
Programmable Gate Arrays ◽  
Computationally Intensive ◽  
Hardware Platforms

Nowadays, computer vision relies heavily on convolutional neural networks (CNNs) to perform complex and accurate tasks. Among them, super-resolution CNNs represent a meaningful example, due to the presence of both convolutional (CONV) and transposed convolutional (TCONV) layers. While the former exploit multiply-and-accumulate (MAC) operations to extract features of interest from incoming feature maps (fmaps), the latter perform MACs to tune the spatial resolution of the received fmaps properly. The ever-growing real-time and low-power requirements of modern computer vision applications represent a stimulus for the research community to investigate the deployment of CNNs on well-suited hardware platforms, such as field programmable gate arrays (FPGAs). FPGAs are widely recognized as valid candidates for trading off computational speed and power consumption, thanks to their flexibility and their capability to also deal with computationally intensive models. In order to reduce the number of operations to be performed, this paper presents a novel hardware-oriented algorithm able to efficiently accelerate both CONVs and TCONVs. The proposed strategy was validated by employing it within a reconfigurable hardware accelerator purposely designed to adapt itself to different operating modes set at run-time. When characterized using the Xilinx XC7K410T FPGA device, the proposed accelerator achieved a throughput of up to 2022.2 GOPS and, in comparison to state-of-the-art competitors, it reached an energy efficiency up to 2.3 times higher, without compromising the overall accuracy.

scholarly journals Fast Modelling of Tsunami Wave Propagation using PC with Hardware Computer Code Acceleration

2001 ◽  
Vol 14 (4) ◽  
pp. 433-444
Author(s):  
Mikhail M. Lavrentiev ◽  
Keyword(s):  
Wave Propagation ◽  
Tsunami Wave ◽  
Hardware Acceleration ◽  
Source Area ◽  
Computer Code ◽  
Analytic Solutions ◽  
Tsunami Source ◽  
Computational Domain ◽  
Field Programmable ◽  
Wave Heights

The field programmable gates array (FPGA) microchip is applied to achieve considerable performance gain in simulation of tsunami wave propagation using personal computer. The two-step Mac-Cormack scheme was used for approximation of the shallow water equations. An idea of PC-based tsunami wave propagation simulation is described. Comparison with the available analytic solutions and numerical results obtained with the reference code show that developed approach provides good accuracy in simulations. It takes less then 1 minute to compute 1 hour of the wave propagation in computational domain that contains 3000 × 2500 nodes. Using the nested greed approach, it is possible to decrease the size of space step from about 300 meters to 10 m. Using the proposed approach, the entire computational process (to calculate the wave propagation from the source area to the coast) takes about 2 min. As an example the distribution of maximal heights of tsunami wave along the coast of the Southern part of Japan is simulated. In particular, the interrelation between maximal wave heights and location of tsunami source is studied. Model sources of size 100 × 200 km have realistic parameters for this region. It was found that only selected parts of the entire coast line are exposed to tsunami wave with dangerous height. However, the occurrence of extreme tsunami wave heights at some of those areas can be attributed to the local bathymetry. The proposed hardware acceleration to compute tsunami wave propagation can be used for rapid (say, during few minutes) evaluation of danger from tsunami wave for a particular location of the coast

scholarly journals Effective coastal boundary conditions for tsunami wave run-up over sloping bathymetry

2014 ◽  
Vol 21 (5) ◽  
pp. 987-1005 ◽  
Author(s):  
W. Kristina ◽  
O. Bokhove ◽  
E. van Groesen
Keyword(s):  
Boundary Condition ◽  
Wave Propagation ◽  
Shallow Water ◽  
Tsunami Wave ◽  
Nonlinear Phenomena ◽  
Tsunami Propagation ◽  
Effective Boundary ◽  
Daunting Task ◽  
Effective Boundary Condition ◽  
Run Up

Abstract. An effective boundary condition (EBC) is introduced as a novel technique for predicting tsunami wave run-up along the coast, and offshore wave reflections. Numerical modeling of tsunami propagation in the coastal zone has been a daunting task, since high accuracy is needed to capture aspects of wave propagation in the shallower areas. For example, there are complicated interactions between incoming and reflected waves due to the bathymetry and intrinsically nonlinear phenomena of wave propagation. If a fixed wall boundary condition is used at a certain shallow depth contour, the reflection properties can be unrealistic. To alleviate this, we explore a so-called effective boundary condition, developed here in one spatial dimension. From the deep ocean to a seaward boundary, i.e., in the simulation area, we model wave propagation numerically over real bathymetry using either the linear dispersive variational Boussinesq or the shallow water equations. We measure the incoming wave at this seaward boundary, and model the wave dynamics towards the shoreline analytically, based on nonlinear shallow water theory over bathymetry with a constant slope. We calculate the run-up heights at the shore and the reflection caused by the slope. The reflected wave is then influxed back into the simulation area using the EBC. The coupling between the numerical and analytic dynamics in the two areas is handled using variational principles, which leads to (approximate) conservation of the overall energy in both areas. We verify our approach in a series of numerical test cases of increasing complexity, including a case akin to tsunami propagation to the coastline at Aceh, Sumatra, Indonesia.

scholarly journals HAL-ASOS Accelerator Model: Evolutive Elasticity by Design

Electronics ◽  
2021 ◽  
Vol 10 (17) ◽  
pp. 2078
Author(s):  
Vítor Silva ◽  
Paulo Pinto ◽  
Paulo Cardoso ◽  
Jorge Cabral ◽  
Adriano Tavares
Keyword(s):  
File System ◽  
Control Unit ◽  
Computing Unit ◽  
System Calls ◽  
Hardware System ◽  
Software Reconfiguration ◽  
Field Programmable ◽  
Significant Performance ◽  
Programmable Gate Arrays ◽  
Virtual File System

To address the integration of software threads and hardware accelerators into the Linux Operating System (OS) programming models, an accelerator architecture is proposed, based on micro-programmable hardware system calls, which fully export these resources into the Linux OS user-space through a design-specific virtual file system. The proposed HAL-ASOS accelerator model is split into a user-defined Hardware Task and a parameterizable Hardware Kernel with three differentiated transfer channels, aiming to explore distinct BUS technology interfaces and promote the accelerator to a first-class computing unit. This paper focuses on the Hardware Kernel and mainly its microcode control unit, which will leverage the elasticity to naturally evolve with Linux OS through key differentiating capabilities of field programmable gate arrays (FPGAs) when compared to the state of the art. To comply with the evolutive nature of Linux OS, or any Hardware Task incremental features, the proposed model generates page-faults signaling runtime errors that are handled at the kernel level as part of the virtual file system runtime. To evaluate the accelerator model’s programmability and its performance, a client-side application based on the AES 128-bit algorithm was implemented. Experiments demonstrate a flexible design approach in terms of hardware and software reconfiguration and significant performance increases consistent with rising processing demands or clock design frequencies.

scholarly journals Effective coastal boundary conditions for tsunami wave run-up over sloping bathymetry

2014 ◽  
Vol 1 (1) ◽  
pp. 317-369
Author(s):  
W. Kristina ◽  
O. Bokhove ◽  
E. van Groesen
Keyword(s):  
Boundary Condition ◽  
Wave Propagation ◽  
Shallow Water ◽  
Tsunami Wave ◽  
Nonlinear Phenomena ◽  
Tsunami Propagation ◽  
Effective Boundary ◽  
Daunting Task ◽  
Effective Boundary Condition ◽  
Run Up

Abstract. An effective boundary condition (EBC) is introduced as a novel technique to predict tsunami wave run-up along the coast and offshore wave reflections. Numerical modeling of tsunami propagation at the coastal zone has been a daunting task since high accuracy is needed to capture aspects of wave propagation in the more shallow areas. For example, there are complicated interactions between incoming and reflected waves due to the bathymetry and intrinsically nonlinear phenomena of wave propagation. If a fixed wall boundary condition is used at a certain shallow depth contour, the reflection properties can be unrealistic. To alleviate this, we explore a so-called effective boundary condition, developed here in one spatial dimension. From the deep ocean to a seaward boundary, i.e., in the simulation area, we model wave propagation numerically over real bathymetry using either the linear dispersive variational Boussinesq or the shallow water equations. We measure the incoming wave at this seaward boundary, and model the wave dynamics towards the shoreline analytically, based on nonlinear shallow water theory over sloping bathymetry. We calculate the run-up heights at the shore and the reflection caused by the slope. The reflected wave is then influxed back into the simulation area using the EBC. The coupling between the numerical and analytic dynamics in the two areas is handled using variational principles, which leads to (approximate) conservation of the overall energy in both areas. We verify our approach in a series of numerical test cases of increasing complexity, including a case akin to tsunami propagation to the coastline at Aceh, Sumatra, Indonesia.

Sign in / Sign up

Export Citation Format

Share Document