Sign In
GTC Logo
GPU
Technology
Conference

March 17-20, 2015 | San Jose, California
Check back often for session updates.
Registration is currently closed. To be notified when registration opens, along with other pertinent news from GTC, sign up here.

Scheduler Planner

Print
Download Pdf
 
  • List View
  • Calender View

 
Refine:
  • Session Levels:
  • |
  • |
  • |
  • |
  • Session Levels:
  •  
  •  
  •  
  •  
  • = Highly Rated Speaker

TALK

Presentation
Details

S5753 - Turbomachinery R&D Acceleration Using Titan

Ravi Srinivasan Aero/Thermodynamic Engineer , Dresser-Rand
Dr. Ravi Srinivasan received his Ph.D. in Aerospace Engineering from Texas A&M University in 2005. His graduate and post-graduate research areas were related to simulation of high-speed flow phenomena. Since joining Dresser-Rand as an Aero/Thermodynamic engineer in 2009, he has focused on the design and CFD modeling of supersonic flow in turbo-machines, including multi-stage and non-linear harmonic modeling.

Dresser-Rand (D-R) is an industrial partner of Oak Ridge Leadership Computing Facility (OLCF) and utilizes the Titan platform to accelerate turbomachinery research and development. In order to take advantage of computing infrastructure at OLCF, D-R has engaged with a third-party CFD software provider to add and modify computational fluid dynamics (CFD) solver modules. The developments include enhancing the scalability of the flow-solver by performing better grid partitioning, implementing GPU based acceleration and significantly improving IO performance. Turbomachinery design at D-R is complemented by employing an optimization process. Titan is the enabling technology that accelerates this process by significantly reducing database generation time and has made it possible to consider implementing optimization as part of R&D. Successful compressor component designs derived from optimization have been experimentally tested by D-R. The steps undertaken for optimization will be presented

Level: Intermediate
Type: Talk
Tags: Supercomputing; Computational Fluid Dynamics; Developer - Performance Optimization

Day: Wednesday, 03/18
Time: 09:30 - 09:55
Location: Room 210D
View Recording
View PDF

S5719 - Hybrid Simulations Using CPU-GPU Paradigm for Reacting Flows

Jeremiah Lee Staff Scientist, United Technologies Research Center
Dr. Jeremiah Lee has 20 years of experience in general area of applied mathematics, computational reactive flow. He has experience in many aspects in this field including, applied mathematics, high performance computing, large scale code development, algorithm development, data reduction, DNS of fundamental flame structures, turbulent flame modeling, acoustics and characteristics in CFD, dynamic chemical kinetics reduction, transition flows, spray dynamics, non-equilibrium multiphase (e.g. superheated fluids) flows, and GPU computing and applications in large scale CFD . His career in applied mathematics started with an investigation of the aerodynamics of baseballs at the Cooper Union. He developed a spectral element based DNS code for reactive flow while he was a graduate student at Princeton. In 1996, he joined the Swiss Federal Institute of Technology (ETH-Z) in Zurich Switzerland where he worked for 6 years as a research staff. In 2002, he returned to the USA and worked at the Combustion Research Facilities at the Sandia National Laboratories. Two years later, he came and joined the combustion group at UTRC. He has 18 publications in refereed journals and has been an invited speaker at the Princeton University, University of Southern California, University of Connecticut, and at the ICDERS.

GPU technology is attractive to computation intensive simulations such as Computational Fluid Dynamics (CFD) of Reacting Flows. A hybrid CPU-GPU paradigm was benchmarked by simulating a canonical CFD problem. A complex turbulent reactive flow was simulated including detailed chemistry that is typically burdensome for CPU based calculations. We achieved 2-5X overall speed-up using CPU-GPU simulations compared to CPU-only simulations. Further details of the CFD problem, hybrid methodology, performance metrics definition and benchmarking results will be presented. This promising technology, if exploited properly, could quickly enable accurate predictions of finite rate chemistry effects, such as pollutant emissions from combustors. ***This talk is part of the "Accelerating Industrial Competitiveness through Extreme-Scale Computing" series chaired by Jack Wells, Director of Science, National Center for Computational Sciences, Oak Ridge National Laboratory.***

Level: Intermediate
Type: Talk
Tags: Supercomputing; Computational Fluid Dynamics; Developer - Performance Optimization

Day: Wednesday, 03/18
Time: 10:30 - 10:55
Location: Room 210D
View Recording
View PDF

S5207 - GiMMiK: Generating Bespoke Matrix-Multiplication Kernels for NVIDIA GPUs

Freddie Witherden Ph.D. Student, Imperial College London
Freddie Witherden
Freddie Witherden studied Physics with Theoretical Physics at Imperial College London between 2008–2012 earning an MSci degree with first class honours. His masters thesis was on the development of a parallel Barnes-Hut type treecode for simulating laser-plasma interactions. Currently, he is a Ph.D. candidate in the department of Aeronautics at Imperial College London under the supervision of Dr Peter Vincent.

Learn how run-time code generation can be used to generate high-performance matrix-multiplication kernels for GPUs. In this talk, I will introduce GiMMiK, an open-source framework for generating bespoke kernels for performing block-by-panel type matrix-matrix multiplications. The techniques employed by GiMMiK will be described in detail. Benchmarks comparing GiMMiK to cuBLAS will be presented and speed-ups of up to 10x will be demonstrated. Specific applications of GiMMiK in the field of high-order computational fluid dynamics will also be highlighted.

Level: Intermediate
Type: Talk
Tags: Computational Fluid Dynamics; Developer - Performance Optimization; Computational Physics; Supercomputing

Day: Thursday, 03/19
Time: 09:00 - 09:25
Location: Room 210B
View Recording
View PDF

S5238 - Multiphysics Simulation Using GPUs

Arman Pazouki Research Associate, University of Wisconsin-Madison
Arman Pazouki
Arman is currently a post-doctoral researcher at the University of Wisconsin-Madison. He received his Ph.D. in Mechanical Engineering and MS in Engineering Mechanics, both from University of Wisconsin-Madison, Madison, WI, MS in Mechanical Engineering from Sharif University of Technology, Iran, and BS in Mechanical Engineering from University of Tehran, Iran. He has more that 10 years of experience of developing code for fluid dynamics and fluid-solid interaction and more than 5 years of experience developing GPU enabled codes using CUDA library.

We present a GPU-based framework for the fully-resolved simulation of interacting rigid and deformable solid objects that move in fluid flow. The fluid dynamics is based on a meshless approach. Moving Lagrangian markers, distributed in the fluid domain as well as on the solid surfaces, are used to capture the fluid dynamics, fluid-solid, and solid-solid interactions. Mass and momentum exchange between neighbor markers are determined in a parallel spatial subdivision algorithm. The solid objects' distributed forces are reduced in parallel via thrust reduction algorithms and used later for temporal update via lightweight GPU kernels. Scenarios containing tens of thousands of floating rigid and flexible objects were exercised on several GPU architectures and the linear scalability was shown.

Level: Advanced
Type: Talk
Tags: Computational Fluid Dynamics; Computational Physics; Developer - Algorithms; Developer - Tools & Libraries

Day: Thursday, 03/19
Time: 09:30 - 09:55
Location: Room 210B
View Recording
View PDF

S5189 - AeroFluidX: A Next Generation GPU-Based CFD Solver for Engineering Applications

Bjoern Landmann Development Engineer, FluiDyna GmbH
Bjoern Landmann is a development engineer at FluiDyna GmbH.

The presentation shows the potential of GPU acceleration for reducing turn-around times of industrial CFD applications. FluiDyna is adressing this issue in a modular approach: the library "Culises" was developed to accelerate matrix operations originating from arbitrary problems. This approach can be complemented by a second module that generates the linear system directly on the GPU – the resulting code being less general, but allowing higher speed-up. The code aeroFluidX is a finite volume solver dedicated to incompressible aerodynamics, combining a SIMPLE algorithm for unstructured grids with state-of-the-art RANS turbulence modelling. MPI-parallelization allows calculations being split-up on multiple GPU-enabled nodes, leading to speed-ups of 2.5-3x for industrial scale problems.

Level: All
Type: Talk
Tags: Computational Fluid Dynamics; Computational Physics; Supercomputing; Automotive

Day: Thursday, 03/19
Time: 10:00 - 10:25
Location: Room 210B
View Recording
View PDF

S5318 - Rolls-Royce Hydra on GPUs Using OP2

Istvan Reguly Postdoctoral Research Associate, University of Oxford
Istvan Reguly
Istvan is a Postdoctoral Research Associate at the University of Oxford, working with Prof. Mike Giles on the acceleration of structured and unstructured mesh computations, collaborating with Rolls-Royce and the Atomic Weapons Establishment in the UK to introduce GPU support and future-proof their codes.

Learn how a Domain Specific Language can be used to accelerate a full-scale industrial CFD application. With OP2, you can easily describe your computational problem at a high level, and then generate CUDA code. We show how parallelization on an unstructured mesh is handled over a cluster of GPUs, and how a range of optimizations can be automatically applied during code generation for GPUs, such as conversion from Array-of-Structures to Structure-of-Arrays, the use of shared memory or caches to improve data reuse. We demonstrate that a 4x performance increase can be achieved with a K40 GPU over a server CPU, and present scaling up to 16 GPUs.

Level: Intermediate
Type: Talk
Tags: Computational Fluid Dynamics; Developer - Programming Languages; Computational Physics

Day: Thursday, 03/19
Time: 10:30 - 10:55
Location: Room 210B
View Recording
View PDF

S5116 - Out-of-Core Proximity Computation on GPU for Particle-Based Fluid Simulations

Duksu Kim Senior researcher, (KISTI) Korea Institute of Science and Technology Information
Duksu Kim
Duksu Kim is currently a senior researcher at KISTI (Korea Institute of Science and Technology Information), South Korea. He received his Ph. D. degree in Computer Science in 2014 from KAIST. His research interests include proximity computation, and large-scale parallel computing.

Lean how to use your GPU for massive-scale particle-based fluid simulations that require a larger amount of memory space than the video memory. We introduce a novel GPU-based neighbor search algorithm used in particle-based fluid simulations such as SPH. With the proposed method, we can efficiently handle a massive-scale particle-based fluid simulation with a limited GPU video memory in out-of-core manner. We have demonstrated that our method robustly handles massive-scale benchmark scenes consisting of up to 65 million particles and requires up to 16 GB memory by using a GPU having only 3 GB memory. It shows up to 26 times higher performance compared to using NVIDIA's mapped memory technique and 51 times higher performance compared to using a CPU core.

Level: Intermediate
Type: Talk
Tags: Computational Fluid Dynamics; Computational Physics; Real-Time Graphics; Developer - Algorithms

Day: Thursday, 03/19
Time: 14:00 - 14:25
Location: Room 210B
View Recording
View PDF

S5756 - Sparse Fluid Simulation in Direct X

Alex Dunn Developer Technology Engineer - Graphics, NVIDIA
As a Developer Technology Engineer for NVIDIA, Alex spends his days passionately working towards advancing real time visual effects in games. A former graduate of, Abertay University's Games Technology Course, Alex got his first taste of graphics programming on the consoles. Now working for NVIDIA his time is spent working on developing cutting edge programming techniques to ensure the highest quality and best player experience possible is achieved.

How to simulate and render game-ready, high resolution fluid in real time on the GPU using DirectX. We'll present a new method for sparsely simulating and rendering traditional grid based fluid systems. By utilizing a simple CPU prediction algorithm, we can update the virtual memory table of the GPU to reflect only the active areas of a simulation volume, providing compressed memory storage and hardware level, memory translation for performing region look ups. This CPU prediction mechanism has a much wider use case than just fluid simulation, and is a must know for anyone planning on using tiled resources in the future.

Level: Advanced
Type: Talk
Tags: Computational Fluid Dynamics; Real-Time Graphics; Developer - Algorithms

Day: Thursday, 03/19
Time: 14:30 - 14:55
Location: Room 210B
View Recording
View PDF

S5304 - Next-Generation CFD: Real-Time Computation and Visualization

Christian Janssen PostDoc, Hamburg University of Technology
Christian Janssen
Christian received his MSc Civil Engineering and his MSc Computational Sciences in Engineering in 2007 and his Ph.D. in Civil Engineering in 2012 from Braunschweig Univ of Tech. In 2011, Christian became PostDoc at the University of Rhode Island and since 2012 has been PostDoc at Hambur University of Tech.

Dive deep into the fascinating world of real-time computational fluid dynamics. We present details of our CUDA-accelerated flow solver for the simulation of non-linear violent flows in marine and coastal engineering. The solver, the efficient lattice boltzmann environment elbe, is accelerated with recent NVIDIA graphics hardware and allows for three-dimensional simulations of complex flows in or near to real-time. Details of the very efficient numerical back end, the pre- and postprocessing tools and the integrated OpenGL visualizer tool will be presented. Join us in this talk to learn about a prototype for next-generation CFD tools for simulation-based design (SBD) and interactive flow field monitoring on commodity hardware.

Level: Advanced
Type: Talk
Tags: Computational Fluid Dynamics; Visualization - In-Situ & Scientific; Computational Physics; Real-Time Graphics

Day: Thursday, 03/19
Time: 15:00 - 15:25
Location: Room 210B
View Recording

S5343 - GPU-Accelerated Fluid Flow with Compute Shaders

Maciej Matyka Assistant Professor, University of Wroclaw
Maciej Matyka
Maciej received his Master Science in Computational Physics, University of Wroclaw, Faculty of Physics and Astronomy, "Numerical investigation of a free surface incompressible flow" (2005), and his Ph.D. in Theoretical Physics, University of Wrocław, Faculty of Physics and Astronomy, "Numerical analysis of the flow through porous media" (2009).

Learn how to utilize compute shaders and write your own, efficient fluid flow solver accelerated with single GPU. First, I will introduce basics of the Lattice Boltzmann method including additional turbulence modelling. Then, an implementation in modern OpenGL will be discussed. I will investigate efficiency of the code and discuss its potential applications in games, medicine and other end-user tools.

Level: Beginner
Type: Talk
Tags: Computational Fluid Dynamics; Visualization - In-Situ & Scientific; Computational Physics

Day: Thursday, 03/19
Time: 15:30 - 15:55
Location: Room 210B
View Recording
View PDF

S5372 - PyFR: Next Generation Computational Fluid Dynamics on GPU Platforms

Freddie Witherden Ph.D. Student , Imperial College London
Freddie Witherden studied Physics with Theoretical Physics at Imperial College London between 2008–2012 earning an MSci degree with first class honours. His masters thesis was on the development of a parallel Barnes-Hut type treecode for simulating laser-plasma interactions. Currently, he is a Ph.D. candidate in the department of Aeronautics at Imperial College London under the supervision of Dr Peter Vincent.

Discover how GPUs are being used to accelerate high-fidelity computational fluid dynamics (CFD) simulations on unstructured grids. In this talk I will (i) introduce the flux reconstruction approach to high-order methods; a discretization that is particularly well-suited to many-core architectures, (ii) introduce our massively parallel implementation PyFR (www.pyfr.org), which through run-time code generation is able to target NVIDIA GPU hardware and, (iii) showcase some of the high-fidelity, unsteady, flow simulations undertaken using PyFR on both desktop and HPC systems.

Level: All
Type: Talk
Tags: Computational Physics; Computational Fluid Dynamics; Developer - Algorithms; Supercomputing

Day: Friday, 03/20
Time: 10:00 - 10:25
Location: Room 210F
View Recording
View PDF

Talk