1、TESLAComputational Fluid Dynamics ModuleComputational Fluid DynamicsGPU Perf compared against Multi-core x86 CPU socket, features and may be a kernel to kernel perf comparisonApplication GPU Features GPU Perf Release Status NotesAltair AcuSolve Linear eqn solver 2x Total Today, release 1.8a FE unstr
2、uctured NS, multi-GPUANSYS Fluent Radiation heat transfer model 10x RHT Model, 2x AMG solver (beta) Today, release 14.5 Multi-GPU RHT model, Single-GPU SolverAutodesk Moldflow Linear eqn solver 1.5x Today, release 2013 FE unstructured NS, single-GPUFluiDyna Culises- OpenFOAM Linear eqn solvers 3x So
3、lver Today, release 1.2 Unstructured NS, single-GPUFluiDyna LBultra LBM, particle CFD 20x Total Today, release 2.0 Structured LBM, multi-GPUVratis SpeedIT- OpenFOAM Solver Linear eqn solvers 6x Solver Today, release 1.2 Unstructured NS, multi-GPUVratis ARAEL Linear eqn solvers 3x Solver Today, relea
4、se 1.0 Single-GPUPrometech Particleworks MPS, particle CFD 4x-9x Total Today, release 3.0 Particle based, multi-GPUSandia NL S3D Chemistry kernel 8x SP, 5x DP kernel Demonstration Structured DNS, multi-GPUSD+ (SU - Jameson) Explicit solver 15x Total In development FE unstructured NS, multi-GPUFEFLO
5、(GMU Lohner) Explicit solver 2-10x Total In development FE unstructured NS, multi-GPUTurbostream Explicit solver 19x Total Today, release 2.0 Structured grid NS, multi-GPUGPU Value to EngineeringComputational Fluid DynamicsSimulation of fluid flow for product developmentWhatWhy Speed of simulations
6、is critical to this workHigher resolution of physics, more complex/real world geometries, better turbulence treatmentHow More problems become practical with GPU accelerationFluiDyna LBultra20x acceleration with 4 GPUs vs. 2 x 6 core CPUs CPU Intel Xeon X5670 2.93 GHz; GPU Tesla M2070GPU READY APPLIC
7、ATIONSAltair AcuSolveAutodesk MoldflowFluiDyna Culises for OpenFOAMFluiDyna LBultraVratis SpeedIT for OpenFOAMPrometech ParticleworksSandia NL and ORNLS3DSD+ (SU-Jameson)FEFLO (GMU-Lohner)TurbostreamANSYS CFD preliminary results of radiation heat transfer view-factor computations on GPUs vs. CPUsRHT
8、 on GPUs will release in 14.0 as betaRadiation HT Applications:NOTE: Growing CPU time of view-factor computations inhibit proper inclusion of radiation HT effectsNOTE: GPU time remains low even as view-factor computations grow very largeANSYS CFD 14.0 Offers First GPU CapabilityUnderhood coolingCabi
9、n comfort HVACFurnace simulationsSolar loads on buildingsCombustor in turbineElectronics passive coolingOther ANSYS CFD Evaluations:Models (e.g. disperse phase)Implicit equation solversOpenFOAM on GPUs ISVs FluiDyna and VratisSpeedIT: 3 GPUs 6x vs. 4 Core i7 CPU Prometech and Particle-based CFD for
10、Multi-GPUsMPS-based method developed at the University of TokyoProf. Koshizuka Results shown for Particleworks 2.5 released in 2011Performance is relative to 4 cores of Intel i7 CPUContact Prometech for license detailshttp:/www.prometech.co.jp Turbostream CFD for Gas Turbine EnginesTurbostream Simul
11、ation Speed-up 19x19xwww.turbostream- | www.many-core.group.cam.ac.uk/ukgpucc2/talks/Brandvik.pdf | Sources: www.hpc.cam.ac.uk/services/darwin.html University of Cambridge DARWIN ClusterCUDA Center of Excellence Since 2008GPU sub-cluster:Dell T5500 servers,32 dual-socket CPUsTesla S1070 GPUs,4 GPUs
12、per socketfor total 128 GPUsSample Turbostream GPU SimulationsTypical Routine SimulationLarge-scale Simulation19x Speeduphttp:/www.turbostream- Turbostream GPU SimulationsTurbostream: CFD for TurbomachineryTokyo Institute of Technology AOKI LaboratoryCFD Research on #5 of Top 500 TSUBAME 2.0Simulati
13、ons that scale to 4000 Fermi GPUsPresentation at Supercomputing 2010 Conference:“Large-scale CFD Applications on TSUBAME 2”Dr. Takayuki Aoki, Global Scientific Information and Computing Center (GSIC) of Tokyo Institute of Technology (Tokyo Tech)http:/ Highlights on CFD Applications From the Top 500F
14、EFLO: Porting of an Edge-Based CFD Solver to GPUsAIAA-2010-0523 Andrew Corrigan, Ph.D., Naval Research Lab; Rainald Lohner, Ph.D., GMUFAST3D: Using GPU on HPC Applications to Satisfy Low Power Computational Requirement AIAA-2010-0524 Gopal Patnaik, Ph.D., US Naval Research LabOVERFLOW: Rotor Wake Mo
15、deling with a Coupled Eulerian and Vortex Particle Method AIAA-2010-0312 Chris Stone, Ph.D., Intelligent LightSOLAR: Unstructured CFD Solver on GPUsJamil Appa, Ph.D., BAE Systems Advanced Technology CentreelsA: Recent Results with elsA on Many-CoresMichel Gazaix and Steve Champagneux, ONERA / Airbus
16、 FranceTurbostream: Turbostream: A CFD Solver for Many-Core ProcessorsTobias Brandvik, Ph.D. , Whittle Lab, University of CambridgeOVERFLOW: Acceleration of a CFD Code with a GPUDennis Jespersen, NASA Ames Research Center48th AIAA Aerospace Sciences Meeting | Jan 2010 | Orlando, FL, USACFD on Future Architectures | Oct 2009 | DLR Braunschweig, DEParallel CFD 2009 | May 2009 | NASA Ames, Moffett Field, CA, USAPublished CFD Developments on Tesla GPU