Intel® HPC Toolkit Release Notes

Where to Find the Release

https://developer.intel.com/oneapi

Also reference the Intel® oneAPI Base Toolkit Release Notes for additional information.

System Requirements

Please see the Intel® HPC Toolkit System Requirements.

Intel® HPC Toolkit Major Component Versions

The following table contains major versions of components in the latest Intel HPC Toolkit 2025.0.1. 

New in Intel® HPC Toolkit 2025.1

Toolkit Level Updates

The Intel HPC Toolkit contains updates included in the 2025.1 update for the Intel oneAPI Base Toolkit, plus the following:

• The Intel® Fortran Compiler now extends its CPU MemorySanitizer support to the device side, including GPUs. This enhancement allows Fortran developers to easily detect and troubleshoot issues in both CPU and device code, ensuring more reliable and robust applications.
• The Intel® Fortran Compiler expands its OpenMP 6.0 standard support by introducing the WORKDISTRIBUTE construct to efficiently distribute work across threads and the INTERCHANGE construct to reorder loops in a loop nest, boosting parallel performance and code optimization.
• The Intel® Fortran Compiler enhances its Fortran 23 support by ensuring consistent kind types for integer arguments in the SYSTEM_CLOCK intrinsic and allowing PUBLIC NAMELIST groups to include PRIVATE variables, providing developers with improved conformance to Fortran 23 language standard and greater code flexibility
• Intel MPI Library applications can now take advantage of new performance tuning for Intel® Xeon® 6 Processors with Performance-Cores (P-Cores), including thread split hand-off for scale-up improvements, CPU inference optimizations for DeepSpeed, and optimizations for point-to-point shared memory operations.

• Intel MPI Library implemented improvements to the default pinning algorithms to improve out-of-the-box resource utilization with benefits for homogeneous and hybrid architecture systems.

• Intel MPI Library now supports device-initiated MPI-RMA functions on supported GPUs in advance of the MPI Standard.
• NEW PRODUCT:  Intel® SHMEM! 
  • Developers have access to a complete Intel SHMEM specification detailing the programming model and supported API calls, with example programs, build and run instructions, and more.
  • Developers using Intel SHMEM can target both device and host with OpenSHMEM 1.5 and 1.6 features including point-to-point Remote Memory Access (RMA), Atomic Memory Operations (AMO), Signaling, Memory Ordering, Teams, Collectives, Synchronization operations, and strided RMA operations.
  • Developers are able to utilize Intel SHMEM and SYCL with API support on device for SYCL work-group and sub-group level extensions of RMA, Signaling, Collective, Memory Ordering, and Synchronization operations and API support on host for SYCL queue ordered RMA, Collective, Signaling, and Synchronization operations.

Intel® oneAPI DPC++ Compiler 2025.1.0

The Intel® oneAPI DPC++/C++ Compiler 2025.1.0

• Added SYCL interoperability with Graphics APIs DirectX* 12 and Vulkan* interop so developers can create highly efficient visual compute, media processing, and rendering applications on Intel® Graphics, enhancing both performance and developer productivity.
• Extends Memory Sanitizer capability to the device side, including GPUs. This enhancement allows to easily detect and troubleshoot issues in both CPU and device code, ensuring more reliable applications.
• Supports ccache* to significantly speed up build times. By caching previous compilations and reusing them, developers can experience faster iterations and more efficient workflows, allowing you to focus on writing high-quality code rather than waiting for builds.
• The code coverage tool now includes GPU support and enhanced CPU coverage for applications using C/C++, SYCL, and OpenMP. It offers detailed analysis and comprehensive HTML reports to identify tested and untested code sections, ultimately improving test coverage and code quality while ensuring easy integration into workflows.

Intel® Fortran Compiler 2025.1

New Fortran 2023 features

• Changes to the SYSTEM_CLOCK intrinsic have been implemented. All integer arguments must have the same kind type, and be of at least default integer kind. Violation of these restrictions is diagnosed as standard messages when option stand f2023 (Windows: /stand:f23)or option standard-semantics appears on the command line.
• PUBLIC NAMELIST groups may now contain PRIVATE variables.

New OpenMP features

• The OpenMP 6.0 WORKDISTRIBUTE construct divides the associated block of statements into units of work that are each executed once by a thread of the enclosing TEAMS construct. A procedure reference in the associated block, with the exception of a reference to MATMUL with fixed-size array arguments, will inhibit the work-sharing properties of this construct at this time.
• The OpenMP 6.0 INTERCHANGE construct reorders loops in a loop nest.

See Intel® Compiler Extension Routines to OpenMP* in the Fortran Developer Guide and Reference for more information.

Intel® MPI Library 2021.15

• Performance Tuning for GNR platform
  • GNR thread split hand-off for scale-up improvements
  • CPU inference optimizations for Deep Speed
  • Optimizations for point to point shared memory operations
• Default Pinning algorithm improvements
  • Pinning for Hybrid architecture
• Functional parity with OMPI on Nvidia GPUs
• Support for device-initiated MPI-RMA (for Intel /NV GPUs) - MPI Standard forward looking feature.
• EFA provider updates
• Bug fixes

New in Intel® HPC Toolkit 2025.0.1

This patch release contains patch release updates for the following components, including bug fixes and quality improvements:

• Intel® oneAPI DPC++ Compiler 2025.0.4
• Intel® MPI Library 2021.14.1

New in Intel® HPC Toolkit 2025.0

Toolkit Level Updates

• Get the most from the latest hardware with new Intel developer tools support for Intel® Xeon® 6 Processors with Performance-Cores (P-Cores), formerly codenamed Granite Rapids and Intel® Core™ Ultra processors (Series 2), formerly codenamed Lunar Lake.
• The Intel® oneAPI HPC Toolkit (HPC Kit) now offers two convenient subset bundles offering smaller downloads for specific developer use cases. Intel® C++ Essentials is for C++ developers is focused on compiling, debugging, and utilizing the most widely used Base Kit performance libraries for Intel CPUs and GPUs. Intel® Fortran Essentials is for Fortran developers focused on compiling, debugging, and utilizing the most widely used HPC Kit performance libraries for Intel CPUs and GPUs.
• The Intel® Fortran Compiler adds new F2023 Standard features, including the AT Edit Descriptor for cleaner output, enhanced string parsing with SPLIT and TOKENIZE functions, and improved numerical precision with upgraded IEEE_ARITHMETIC capabilities, all designed to optimize your coding efficiency and application performance.
• Maximize your Fortran application's parallel processing capabilities with our latest Intel® Fortran Compiler release, featuring OpenMP 6.0 enhancements such as conditional TEAMS construct execution with the new IF clause, flexible TARGET constructs with DEVICE_TYPE clauses, and enhanced device targeting and task affinity control with OpenMP 5.1 updates, all designed to give developers greater control and efficiency in high-performance computing environments.
• Unlock advanced parallelism in your Fortran applications with Intel® Fortran compiler's latest enhancements, now supporting arrays of coarrays and the ability to create allocatable arrays with coarray components, offering developers dynamic, high-performance data structures for sophisticated coarray programming.
• Achieve high scale out and scale up performance on Intel® Xeon® 6 Processors with Intel® MPI Library including P-core pinning for optimized balancing of asymmetric CPU topologies.
• Intel MPI Library now offers a full MPI 4.0 implementation including Partitioned Communication, Improved Error handling, and Fortran 2008 support
• New Intel MPI Library optimizations for MPI_Allreduce improve scale up and scale out performance for Intel GPUs
• ISO C++ Parallel STL code runs on CPU and offloads to GPU using Intel® oneAPI DPC++/C++ Compiler.
• Experience dynamic and flexible GPU programming with Intel® oneAPI DPC++/C++ Compiler's SYCL Bindless Textures support, utilizing textures at runtime without compile-time knowledge for improved performance and scalability in C++ with SYCL applications, alongside powerful new LLVM sanitizers to streamline development and ensure enhanced device code reliability.
• Maximize your application's efficiency with Intel oneAPI DPC++/C++ Compiler's performance optimization features, tailored for the latest Intel platforms including Intel® Xeon® 6 Processors and Intel® Core™ Ultra processors (Series 2), to deliver peak performance and cutting-edge computing experiences.
• Leverage enhanced OpenMP standards support and performance enhancements with the Intel® oneAPI DPC++/C++ Compiler, including OpenMP 5.x and 6.0 features for increased efficiency and flexibility, complemented by upgraded compiler opt-report capabilities for in-depth performance insights and optimization feedback.
• GPU kernels run faster with Intel® oneAPI DPC++ Library (oneDPL) improved performance by up-to 4X for algorithms including reduce, scan and many other functions. 
• Use oneDPL Range-based algorithms with over 20 new C++20 standard ranges and views to accelerate on multiarchitecture devices.
• Intel® Math Kernel Library (oneMKL)SYCL Discrete Fourier Transform API is easier to use and to debug with key compilation messages added for type safety, reducing time to develop your application, especially when targeting Intel GPUs.
• HPC workloads using oneMKL single precision 3D real in-place FFTs run faster on Intel® Data Center GPU Max Series.
• Multi-threads apps run faster with Intel® oneAPI Threading Building Blocks (oneTBB) task_group, flow_graph and parallel_for_each improved scalibility
• Get result faster using oneTBB flow graph to process overlapping messages on a shared graph, waiting for a specific message using the new try_put_and_wait experimental API
• Intel® Integrated Performance Primitives (Intel® IPP) now boasts CET-enabled protection, safeguarding your software against control-flow attacks and mitigates exploitation risks. Safeguard your software with cutting-edge, hardware-enforced security measures.
• Use Intel® Cryptography Primitives Library to turbocharge RSA encryption (2K, 3K, 4K) with multi-buffer capabilities—achieving up to 4x the speed of OpenSSL.
• Use Intel® Cryptography Primitives Library to dive into the future of hashing with our enhanced SM3 algorithm, now 5x faster thanks to the SM3_NI instructions
• Intel® VTune™ Profiler now identifies GPU-bound bottlenecks, optimize rendering pipelines, and improve overall application responsiveness for media and content creation applications on Intel® Core™ Ultra 200V, codenamed Lunar Lake.
• Intel® VTune™ Profiler now identifies and optimizes device-side inefficiencies for Direct X APIs.
• Intel® Advisor introduces a more adaptable kernel matching mechanism, enabling developers to identify and analyze code regions relevant to their specific optimization goals. The integration with the XCG app streamlines the process of offloading computation to GPUs, enhancing performance on Intel's latest hardware.
• Intel® oneAPI Deep Neural Network Library (oneDNN) dramatically boosts performance for Large Language Models and scaled dot-product subgraphs.
• Intel® oneAPI Communications Library (oneCCL) now includes optimizations that enable workloads to scale and perform even better than before. Important enhancements have been made to key collectives, and even more optimizations are now available on single-node CPU configurations.  
• Save time with Intel® DPC++ Compatibility Tool to easily migrate your CUDA code and CMake build script to SYCL as demonstrated by auto migration of more APIs used by popular AI, HPC and rendering apps.  The migrated code is easy to comprehend with SYCLcompat, easy to debug using CodePin, and runs performantly on Nvidia GPUs
• Free your imaging apps from vendor lock-in using Intel® DPC++ Compatibility Tool to migrate bindless textures APIs to SYCL image extension
• Intel® Distribution for GDB* adds support for Intel® Core™ Ultra processors (Series 2) on Windows* allowing developers to efficiently debug application code on these new CPUs and GPUs.
• Intel® Distribution for GDB* rebases to GDB* 15 staying current and aligned with the latest enhancements supporting effective application debug.
• Intel® Distribution for GDB* enhances the developer experience, both on the command line and when using Microsoft* Visual Studio and Visual Studio Code* by boosting the debugger performance and refining the user interface.

Intel® oneAPI DPC++/C++ Compiler 2025.0

• New Features for standard support such as OpenMP 6.0, Fortran 2023 and SYCL 2020.
• Optimized for new Intel hardware including EMR, GNR, BMG, and LNL, with features such as cache hints and new data types for AI applications, delivering improved efficiency and computing power.
• Performance tuning for AI framework and HPC applications.
• Improvements such opt-report enhancement for better user experiences.
• Bindless Textures Support implemented for Intel GPUs (DG2, Arc).
• Device code now supports sanitizers to help identify issues.
• New SYCL Offload model introduced with --offload-new-driver.
• Performance enhancements and early support for OpenMP 6.0 features.
• OpenMP loop rotation issue fixed
• New OpenMP 6.0 DEVICE_TYPE clause for TARGET construct 
• Mandatory OpenMP offload support enabled
• Improved user experience with additional optimization report information for OpenMP offloading.
• SYCL now offers functionality comparable to CUDA textures for Intel Client GPUs.
• Advanced support for optional kernel features in Ahead-Of-Time (AOT) compilation mode.
• SYCL ABI has undergone breaking changes.

Intel® Fortran Compiler 2025.0

• New F2023 Standard features for the runtime library including the AT Edit Descriptor to trim trailing blanks before output, and SPLIT and TOKENIZE intrinsic functions for parsing character strings. IEEE_ARITHMETIC has been enhanced with F2023 IEEE_MAX, IEEE_MAX_MAG, IEEE_MIN, IEEE_MIN_MAG, along with F2023 behavior changes for IEEE_MAX_NUM, IEEE_MAX_NUM_MAG, IEEE_MIN_NUM and IEEE_MIN_NUM_MAG
• New OpenMP features in this release:
  1. OpenMP 6.0 An IF clause allowed on TEAMS construct
  2. OpenMP 6.0 A DEVICE_TYPE clause may now appear on a TARGET construct.
  3. OpenMP 5.1 The DEVICE_TYPE clause may appear on a DECLARE TARGET directive.
  4. OpenMp 5.1 AFFINITY clause  is now permitted on a TASK directive.
• New work on Coarrays that include arrays of coarrays, a data object with a coarray component may now be an array and may be allocatable.
• The standard-semantics option is now sensitive to a standards conformance level set by the stand option.  If no stand option is specified, standard semantics sets options to conform to Fortran 2018 standard behavior.
• The Intel® Fortran Developer Guide and Reference received extensive updates to remove older material, refresh existing examples, and added supported Fortran 2018 and 2023 Fortran Language features

Intel® MPI Library 2021.14.0

• GNR /SRF Tuning and optimizations for both scale out and scale up.
  • GNR - Improved CPU pinning library for optimized balancing for asymmetric CPU topologies. (IPL2 becomes PoR)
• MPI 4.0 compliance- Support for Partition communication, Improved Error handling, Fortran 2008 support
• MPI_Allreduce scale-up and scale-out optimizations for Intel GPUs
• Intel GPU aware reduce optimizations
• Windows IFX support for MPI
• OFI and provider update to latest open-source versions

Deprecation Notice

• Intel® Fortran Compiler Classic (ifort) is now discontinued in oneAPI 2025 release.

Installation Instructions

Please see the Installation Guide for oneAPI Toolkit.

Known Issues

• When using Intel® oneAPI DPC++/C++ Compiler on a Linux machine, users may run into an issue if the highest version of GNU gcc detected doesn't have the equivalent g++ package installed. More details on the error and workarounds can be found Fatal Error: 'C++ Header' File Not Found with Intel® oneAPI DPC++/C++ Compiler.
• Please read the whitepaper on Challenges, tips, and known issues when debugging heterogeneous programs using DPC++ or OpenMP offload.
• Running any GPU code on a Virtual Machine is not supported at this time.
• If you have chosen to download the Get Started Guide to use offline, viewing it in Chrome may cause the text to disappear when the browser window is resized. To fix this problem, resize your browser window again, or use a different browser.
• Eclipse* 4.12: the code sample project created by IDE plugin from Makefile will not build. It is a known issue with Eclipse 4.12. Please use Eclipse 4.9, 4.10 or 4.11.

Deprecated/Removed Features

• Intel® Fortran Compiler Classic (ifort) is now discontinued in oneAPI 2025 release.
• The Diagnostics Utility for oneAPI is now deprecated and will be discontinued with the 2025.0 release.
• Intel® Trace Analyzer and Collector is no longer included in the Intel® HPC Toolkit as of update 2024.1.0. It is now downloadable as a standalone package and it will be discontinued in 2025 or later.
• Intel® Inspector is no longer included in the Intel® HPC Toolkit as of update 2024.1.0. It is now downloadable as a standalone package and it will be discontinued in 2025 or later.
• Intel® Fortran Compiler Classic (ifort) is now deprecated and will be discontinued in late 2024. Intel recommends that customers transition now to using the LLVM-based Intel® Fortran Compiler (ifx) for continued Windows* and Linux* support, new language support, new language features, and optimizations.
  For more information on ifx, see the Intel® Fortran Compiler Developer Guide and Reference and the Porting Guide for ifort Users to ifx.
• The following OS support is now deprecated and will be discontinued with our 2025.0 release.
  • CPU:
    • SUSE Linux Enterprise Server (SLES) version 15 SP3
    • Ubuntu Linux version 20.04
    • Fedora Linux version 37
    • Debian Linux version 11
    • Amazon Linux version 2022
  • GPU:
    • Red Hat Enterprise Linux (RHEL) version 8.6

Tools Included in Intel® HPC Toolkit

• Intel® oneAPI DPC++/C++ Compiler Pro - Release Notes
• Intel® Fortran Compiler - Release Notes
• Intel® MPI Library - Release Notes

Previous oneAPI Releases

• Intel® HPC Toolkit 2024
  • Release Notes
  • System Requirements
• Intel® oneAPI HPC Toolkit 2022
  • Release Notes
  • System Requirements
• Intel® oneAPI HPC Toolkit 2021
  • Release Notes
  • System Requirements

Notices and Disclaimers

Intel technologies may require enabled hardware, software or service activation.

No product or component can be absolutely secure.

Your costs and results may vary.

© Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.

No license (express or implied, by estoppel or otherwise) to any intellectual property rights is granted by this document.

The products described may contain design defects or errors known as errata which may cause the product to deviate from published specifications. Current characterized errata are available on request.

Intel disclaims all express and implied warranties, including without limitation, the implied warranties of merchantability, fitness for a particular purpose, and non-infringement, as well as any warranty arising from course of performance, course of dealing, or usage in trade.