环境科学领域应用和优化.pptx

环境科学领域应用与优化NWS applications that require petascale computingSource: Michael B?ttinger / DKRZImpact of increased resolution is exponentialIncreasing ComplexitySource: Rich Loft, NCAR; presented at CAS2K7Intel SW tools for HPCIntel? Parallel Studio XE Design, build, verify, and tune为程序设计，代码构建，校验和调试提供全方位支持C++, C, Fortran*, Python* and Java*支持C++, C, Fortran and Java* 等编程环境Standards-driven parallel models: OpenMP*, MPI, and TBB支持标准的并行模型Highlights from 2017 editionFaster Python application performance using Intel? Distribution for Python and Intel? VTune? Amplifier XE.使用Intel集成的 Python和 Intel? VTune? Amplifier XE 调优工具，提升Python 应用的性能Faster deep learning on Intel? architecture using Intel? Math Kernel Library and Intel? Data Analytics Acceleration Library使用Intel MKL/DAAL类库提升基于IA平台的深度学习系统性能Quickly assess application performance using snapshot features of Intel? VTune? Amplifier XE and Intel? Trace Analyzer and Collector使用Intel vTune快照功能和追踪收集分析工具快速评价应用性能 Scale to next-generation platforms including the latest Intel? Xeon Phi? processor. Optimizations for Intel? AVX-512, high bandwidth memory, and explicit vectorization for compiler and analysis tools.包括Xeon Phi在内的IA新一代平台的部署，Intel编译器和解析工具帮助开发者实现AVX-512，高带宽内存和向量化的优化http://intel.ly/perf-toolshttphttp://intel.ly/perf-tools://intel.ly/perf-toolsIntel? Parallel Studio XE Intel? InspectorMemory and Threading CheckingIntel? AdvisorVectorization Optimization and Thread PrototypingProfiling, Analysis, and ArchitectureIntel? Cluster CheckerCluster Diagnostic Expert SystemIntel? VTune? AmplifierPerformance ProfilerIntel? Trace Analyzer and CollectorMPI ProfilerCluster ToolsIntel? MPI LibraryIntel? Data Analytics Acceleration LibraryOptimized for Data Analytics Machine LearningIntel? Integrated Performance PrimitivesImage, Signal, and Compression RoutinesPerformance LibrariesIntel? Math Kernel LibraryOptimized Routines for Science, Engineering, and FinancialIntel? Threading Building BlocksTask-Based Parallel C++ Template LibraryIntel