- Speed up applications with industry-leading, standards-based C and C++ tools.
- Experience seamless compatibility with popular compilers, development environments, and operating systems.
- Get superior vectorization and parallelization capabilities that include Intel® Advanced Vector Extensions 512 (Intel® AVX-512) instructions, using the latest OpenMP* 5.0 parallel programming model.
- Gain superior Fortran application performance.
- Get extensive support for the latest Fortran standards (including full Fortran 2008 and expanded Fortran 2018), with backwards compatibility to Fortran 77.
- Boost Single Instruction Multiple Data (SIMD) vectorization and threading capabilities (including Intel AVX-512 instructions) using the latest OpenMP parallel programming model.
- Accelerate Python* application performance with NumPy, SciPy, and scikit-learn*—all optimized for Intel® processors.
- Leverage the power of native performance libraries, multithreading, and the latest vectorization instructions for faster computing. Scale with mpi4py. Supports Python 3.x.
- Accelerate math processing routines with the fastest, most-used math library for Intel® and compatible processors.
- Maximize performance with highly optimized, threaded, and vectorized math functions that scale on current and future Intel® platforms.
- Use de facto standard APIs for simple code integration.
- Includes highly optimized machine learning and analytics functions.
- Simultaneously ingests data and computes results for the highest throughput performance.
- Supports batch, streaming, and distributed usage models to meet a range of application needs.
- Deliver highly optimized image and signal processing, data compression, and cryptography applications using Intel® Streaming SIMD Extensions 2, 3, and 4, and Intel® Advanced Vector Extensions (Intel® AVX) instruction sets.
- Multicore, multi-OS, and multiplatform ready. Plug in and use APIs to quickly improve applications.
- Reduces development time and costs.
- Specify tasks instead of manipulating threads. Map your logical tasks onto threads with full support for nested parallelism.
- Load balance and cut task execution time with proven, efficient parallel patterns, and work stealing.
- Use open source and licensed versions for Linux*, Windows*, macOS*, and Android* that are compatible with multiple compilers and Intel processors.
- Simplify threading parallelism with a flow graph feature that expresses dependency and data flow graphs intuitively and easily.
- Available for C++ only.
英特尔的编译器针对非英特尔微处理器的优化程度可能与英特尔微处理器相同（或不同）。这些优化包括 SSE2、SSE3 和 SSSE3 指令集和其他优化。对于在非英特尔制造的微处理器上进行的优化，英特尔不对相应的可用性、功能或有效性提供担保。该产品中依赖于微处理器的优化仅适用于英特尔微处理器。某些非特定于英特尔微架构的优化保留用于英特尔微处理器。关于此通知涵盖的特定指令集的更多信息，请参阅适用产品的用户指南和参考指南。