首页| 论坛| 消息
主题:Intel OneAPI 2022.3.0
pony8000发表于 2022-09-28 20:50
Intel OneAPI 2022.3.0 | 12.5 Gb
The oneAPI Toolkits development team is pleased to announce the availability of Intel oneAPI Base & HPC Toolkit 2022.3 is a comprehensive suite of development tools that make it fast and easy to build modern code that gets every last ounce of performance out of the newest Intel processors in high-performance computing (HPC) platforms.

New in Intel oneAPI Base Toolkit 2022.3

Toolkit Level Updates
- Intel VTune Profiler 2022.4 , Intel Advisor 2022.3, Intel Inspector 2022.3 and Intel Trace Analyzer and Collector 2021.7 may not include all the latest functional and security updates. New versions of Intel® VTune™ Profiler, Intel® Advisor, Intel® Inspector and Intel® Trace Analyzer and Collector are targeted to be released in December 2022 and will include additional functional and security updates. Customers should update to the latest version as it becomes available.
Intel oneAPI DPC++ Compiler
- Intel oneAPI DPC++/C++ Compiler adds more SYCL* 2020 features to improve developer productivity for programming various hardware accelerators including GPUs and FPGAs.
- The compiler enhances OpenMP* 5.x compliance and improves performance.
- Intel C++ Compiler Classic (icc) is deprecated and will be removed in a oneAPI release in the second half of 2023. Intel recommends that customers transition now to using the LLVM-based Intel® oneAPI DPC++/C++ Compiler (icx) for continued Windows* and Linux* support, new language support, new language features, and optimizations. Note that starting with release 2021.7, macOS* support is limited to Mac* computers with Intel® Processors. For more information on icx, see the Intel® oneAPI DPC++/C++ Compiler Developer Guide and Reference.
Intel DPC++ Compatibility Tool
- The Intel DPC++ Compatibility Tools aded support for CUDA 11.7 header files.
- This release also adds migration of CUDA runtimes and driver APIs, including cuDNN APIs, NCCL APIs, Thrust APIs, cuBLAS APIs, cuFFT APIs.
- User defined migration rules are extended to support include statement, enum type, class type, getter and setter in user defined class rule, templates in type rule and namespaces in API rule.
Intel oneAPI DPC++ Library
- Expanded support of the C++ standard library in SYCL kernels with nine additional heap and sorting algorithms.
Intel oneAPI Math Kernel Library
- Cost efficient GPU resource allocation - Idle cycles lower GPU cost efficiency, but the latest BLAS GPU device timing support and GPU VERBOSE features help get insights on GPU utilization and further optimization.
- Faster recovery -The improved argument checks and exception handling for BLAS DPC++ interface ensures faster and easier detection of exceptions(such as runtime threading errors) and quicker recovery.
- Improved portability & compatibility - Extended OpenMP* cluster offload capability supports latest OpenMP* 5.1 directive syntax for LAPACK (with Intel® Compilers for Linux* OS).
Intel oneAPI Threading Building Blocks
- This release addresses bugs to improve performance and quality.
Intel oneAPI Video Processing Library
- Intel oneAPI Video Processing Library now includes the ability to provide extensive data about what is encoded, allowing applications to dynamically adjust their approach based on immediate feedback, opening up opportunities for quality improvement and algorithm innovation.
- Feature enhancements to sample transcoding allow for easier evaluation of GPU real time features such as HDR, pipeline latency, and AV1 temporal layers.
- New low latency Linux initialization for the Intel Data Center Flex Series GPU means that customers can start processing video on multi-GPU systems in a few milliseconds.
Intel Distribution for GDB*
- Intel Distribution for GDB* improves the developer experience for stepping though code when using the Fortran Expression Evaluator within Microsoft* Visual Studio.
- The debugger provides enhanced usability and stability for seamless GPU side debugging.
Intel Distribution for Python
- The numba_dppy Python module has been renamed to numba_dpex, requiring a change of “import numba_dppy” to “import numba_dpex”.
- CVE and bug fixes.
Intel VTune Profiler
- Intel VTune Profiler has been updated to include recent versions of 3rd party components, which include functional and security updates.
Intel Advisor
- Intel Advisor has been updated to include recent versions of 3rd party components, which include functional and security updates.

New in Intel oneAPI HPC Toolkit 2022.3

Toolkit Level Updates
- Intel VTune Profiler 2022.4 , Intel® Advisor 2022.3, Intel Inspector 2022.3 and Intel® Trace Analyzer and Collector 2021.7 may not include all the latest functional and security updates. New versions of Intel® VTune™ Profiler, Intel® Advisor, Intel® Inspector and Intel® Trace Analyzer and Collector are targeted to be released in December 2022 and will include additional functional and security updates. Customers should update to the latest version as it becomes available.
Intel DPC++/C++ Compiler Classic
- Intel oneAPI DPC++/C++ Compiler adds more SYCL* 2020 features to improve developer productivity for programming various hardware accelerators including GPUs and FPGAs.
- The compiler enhances OpenMP* 5.x compliance and improves performance.
- The Intel C++ Compiler Classic has been updated to include recent versions of 3rd party components, which include functional and security updates.
- Intel C++ Compiler Classic (icc) is deprecated and will be removed in a oneAPI release in the second half of 2023. Intel recommends that customers transition now to using the LLVM-based Intel® oneAPI DPC++/C++ Compiler (icx) for continued Windows* and Linux* support, new language support, new l

浏览大图

浏览大图

浏览大图
附件: 地址.txt (1 K) 下载次数:3 该附件被加密,需要89以上积分才能下载
下一页 (1/4)
回帖(11):
11楼:
10楼:谢谢,经典资料,正在找
9楼:经典资料哈,这都能找到

全部回帖(11)»
最新回帖
收藏本帖
发新帖