Intel oneAPI

Intel oneAPI is a unified, open programming model designed by Intel to simplify development across diverse hardware architectures including CPUs, GPUs, FPGAs, and other accelerators. It provides a comprehensive suite of compilers, libraries, analysis tools, and migration utilities optimized for high-performance computing (HPC), artificial intelligence (AI), and data analytics.

Key benefits of Intel oneAPI in HPC environments include:

• Cross-architecture support: Write code once and run it on multiple hardware platforms.

• Performance optimization: Libraries and tools are tuned for Intel hardware.

• Open standards: Based on SYCL and DPC++, avoiding vendor lock-in.

• Interoperability: Compatible with existing models like MPI, OpenMP, and CUDA.

The Official Documentation Links

• Intel oneAPI Overview

• Intel Compilers

• Intel MKL

• Intel MPI

• Intel VTune Profiler

• Intel Advisor

• Intel DPC++ Compatibility Tool

• Intel DNNL

• Intel TBB

• Intel IPP

• Intel CCL

The Intel OneAPI modules on Aire

This page documents all Intel oneAPI modules available on the HPC system, including descriptions, official links, usage instructions, and licensing notes.

Version	Module Load Command	Description
2025.0.4	module load intel/oneapi/compiler/2025.0.4	Intel C/C++/Fortran compilers
2025.0.4	module load intel/oneapi/compiler-intel-llvm/2025.0.4	LLVM-based compiler with SYCL/DPC++ support
2025.0.4	module load intel/oneapi/compiler-rt/2025.0.4	Runtime libraries for compiled applications
2025.0	module load intel/oneapi/mkl/2025.0	Math Kernel Library for linear algebra, FFTs, etc.
2021.14	module load intel/oneapi/mpi/2021.14	Intel MPI implementation
2025.0	module load intel/oneapi/vtune/2025.0	Advanced performance profiler
2025.0	module load intel/oneapi/advisor/2025.0	Performance analysis and optimization tool
2025.0.0	module load intel/oneapi/debugger/2025.0.0	Debugging tool for DPC++ and other languages
2025.0.0	module load intel/oneapi/dpct/2025.0.0	CUDA-to-SYCL migration tool
3.6.1	module load intel/oneapi/dnnl/3.6.1	Deep Neural Network Library
2022.7	module load intel/oneapi/dpl/2022.7	Data Parallel STL-like algorithms
2022.0	module load intel/oneapi/tbb/2022.0	Threading Building Blocks
2021.14.0	module load intel/oneapi/ccl/2021.14.0	Collective Communications Library
2025.0.0	module load intel/oneapi/dev-utilities/2025.0.0	Developer utilities
0.9.1	module load intel/oneapi/umf/0.9.1	Unified Memory Framework
2022.0	module load intel/oneapi/intel_ipp_intel64/2022.0	Signal/image processing primitives
2025.0	module load intel/oneapi/intel_ippcp_intel64/2025.0	Cryptographic primitives

Licensing

Intel oneAPI toolkits are generally free to use and do not require a license for most components. However, some tools and libraries may have specific licensing terms. For details, refer to:

• Intel oneAPI Licensing FAQ

Note

Previous versions of Intel oneAPI included both ifort and ifx compilers. However, starting from the 2025 version, only ifx is supported.