VIP20
本课程属于：007-Advanced CAE Simulation and Engineering Analysis English Technical Courses · 共 20 门相关课程，组合学习更完整。
查看所属VIP全部课程 →
← 返回所属目录 咨询 / 购买

High-Performance Computing for Large CAE Simulations

30 Chapters
View Course Samples Online
01
HPC-driven CAE simulation landscape
large-scale structural, fluid, thermal, and multiphysics workloads, industrial drivers, and computational bottlenecks.
HPC-drivenCAE
02
Parallel computing foundations for CAE
shared-memory and distributed-memory models, Amdahl's law, Gustafson's law, and scalability metrics in simulation practice.
Parallelcomputing
03
CAE solver architectures
implicit and explicit time integration, sparse linear systems, nonlinear iteration, and domain decomposition strategies for large models.
CAEarchitectures
04
Hardware platforms for large CAE
multicore CPUs, GPUs, many-core accelerators, interconnects, memory hierarchies, and storage subsystems.
Hardwareplatforms
05
Profiling and performance analysis of CAE codes
timers, hardware counters, roofline modeling, communication tracing, and hotspot identification.
Profilingperformance
06
Shared-memory parallelization with OpenMP
thread teams, scheduling, data scoping, reductions, and tuning for assembly and element loops.
Shared-memoryparallelization
07
Distributed-memory parallelization with MPI
point-to-point and collective communication, communicators, nonblocking operations, and CAE-specific patterns.
Distributed-memoryparallelization
08
Hybrid MPI+OpenMP parallelization
thread affinity, communication overlap, task granularity, and performance trade-offs in large simulations.
MPI+OpenMPparallelization
09
GPU acceleration for CAE
CUDA and OpenMP target offload, memory coalescing, kernel design, and porting element and solver kernels.
GPUacceleration
10
Sparse matrix assembly and storage on HPC systems
CSR, ELL, block formats, parallel assembly, and memory bandwidth considerations.
assemblyHPC
11
Parallel sparse linear solvers
direct factorization, iterative Krylov methods, preconditioners, and scalability for large CAE systems.
Parallelfactorization
12
Domain decomposition methods for CAE
mesh partitioning, METIS and ParMETIS, load balancing, halo exchange, and interface handling.
decompositionCAE
13
Parallel nonlinear solvers
Newton-Krylov-Schwarz methods, line search, convergence checks, and robustness for large deformation and contact.
Parallelnonlinear
14
Time-domain parallelism and transient CAE
parallel-in-time methods, checkpointing, and scheduling for long-duration simulations.
Time-domainparallelism
15
I/O and data management for large CAE
parallel file systems, MPI-IO, HDF5, ADIOS, in-situ visualization, and checkpoint-restart strategies.
I/Omanagement
16
Coupled multiphysics HPC workflows
partitioned and monolithic coupling, inter-code communication, load balancing, and co-simulation.
multiphysicsHPC
17
Mesh generation and adaptation at scale
parallel meshing, octree and advancing-front methods, dynamic load balancing, and error-driven refinement.
generationadaptation
18
Uncertainty quantification and design exploration on HPC
sampling, surrogate models, parallel ensembles, and sensitivity analysis for CAE.
Uncertaintyquantification
19
Optimization under HPC for CAE
gradient-based and derivative-free methods, adjoint solvers, and parallel Pareto front exploration.
OptimizationHPC
20
Performance portability and code modernization
Kokkos, RAJA, SYCL, OpenACC, and refactoring legacy CAE solvers for heterogeneous systems.
Performanceportability
21
Containerization and reproducibility for CAE HPC
Singularity and Apptainer, environment modules, build systems, and portable simulation stacks.
Containerizationreproducibility
22
Job scheduling and resource management
Slurm, PBS, and LSF, job arrays, topology-aware placement, and queue policies for CAE workloads.
schedulingresource
23
Cloud and cluster HPC for CAE
elastic scaling, data staging, cost-performance trade-offs, and hybrid on-premises/cloud simulation pipelines.
HPCCAE
24
Benchmarking and validation of large CAE simulations
representative test cases, weak and strong scaling studies, and performance regression tracking.
Benchmarkingvalidation
25
Fault tolerance and resilience in long CAE runs
checkpointing, migration, silent data corruption detection, and recovery workflows.
toleranceresilience
26
AI-enhanced HPC for CAE
physics-informed neural networks, operator learning, surrogate acceleration, and hybrid solver-AI coupling.
AI-enhancedHPC
27
Exascale-ready CAE
extreme-scale challenges, mixed precision, communication-avoiding algorithms, and next-generation solver design.
Exascale-readyCAE
28
Case study
large-scale structural analysis on a leadership-class system, from model preparation to parallel solution and post-processing.
large-scalestructural
29
Case study
high-fidelity CFD simulation using GPU-accelerated solvers, including meshing, partitioning, and performance tuning.
high-fidelityCFD
30
Capstone project
end-to-end HPC workflow for a large CAE problem, covering parallelization, profiling, optimization, and scalability reporting.
Capstoneend-to-end