Intel oneAPI 2022 toolkits help developers improve productivity and accelerate innovation

Intel has released its oneAPI 2022 toolkits, expanding cross-architecture functionality to provide developers with greater utility and architectural choice.

“I am impressed with the breadth of over 900 technical enhancements made by the oneAPI software engineering team to accelerate development time and performance of critical application workloads on client and server processors and GPUs. from Intel. The rich set of oneAPI technologies comply with key industry standards, with deep technical innovations that enable application developers to achieve the best possible runtime performance from the cloud to the edge. Multilingual support and performance acceleration across architectures are ready today in our oneAPI 2022 release to further enable programmer productivity on Intel platforms, ”said Greg Lavender, Chief Technology Officer, Intel , senior vice president and general manager of the Software and Advanced Technology group. .

New features include the world’s first unified compiler implementing C ++, SYCL, and Fortran, data-parallel Python for CPUs and GPUs, advanced accelerator performance modeling and tuning, and performance acceleration for workloads. AI visualization and ray tracing. The oneAPI inter-architecture programming model provides developers with tools to improve productivity and speed of code development when building inter-architecture applications.

According to an Evans Data survey, 40% of developers target heterogeneous systems that use more than one type of processor, processor core, or coprocessor. The pace of innovation is only accelerating thanks to deep cross-architecture computing with an API on heterogeneous structures of CPUs, GPUs, FPGAs and other accelerators, all of which require software to tie these together in a solution. complete. With oneAPI, developers have the freedom to choose the best hardware for a given solution without the economic and technical burdens of proprietary programming models.

Intel oneAPI 2022 Toolkits deliver performance and productivity with a comprehensive set of advanced tools, including compilers, libraries, pre-optimized frameworks, analyzers and debuggers. There are over 900 new and improved features added over the past year that reinforce each tool in fundamental and domain-specific toolkits. They can now be downloaded or used for free in Intel DevCloud. Highlights include:

Inter-architecture programming

  • Intel created the world’s first unified compiler implementing C ++, SYCL, and Fortran for CPUs and GPUs using a common LLVM backend.
  • Accelerated computing on CPUs and GPUs for Python, today’s most popular programming language.
  • The Intel DPC ++ Compatibility Tool has been enhanced to automatically migrate 90% to 95% of CUDA code to SYCL / DPC ++.

Performance on the latest hardware

  • Hardware support – Intel oneAPI Toolkits are optimized to enable advanced features of the latest and upcoming new hardware including 12th Generation Intel Core processors with AVX-VNNI, Next Generation Intel Xeon Scalable processors named Sapphire Rapids with Intel Advanced Matrix Extension (Intel AMX), and upcoming GPUs for Xe customers and data centers.

AI performance optimizations

  • The performance of the deep learning framework is accelerated up to 10 times compared to previous versions with the latest Intel optimizations for TensorFlow and Intel Optimization for PyTorch.
  • The new Intel extension for Scikit-learn accelerates machine learning algorithms more than 100 times on Intel processors compared to the original open source version.
  • Introduced Intel Neural Compressor to achieve increased inference performance through post-training optimization techniques on multiple deep learning frameworks.

Advanced tools for development productivity

  • Expert performance analysis of processors and accelerators – Intel VTune Profiler flame graph display helps enhance the ability to visualize performance hot spots. Intel Advisor’s accelerator performance modeling allows developers to estimate the performance benefits of offloading to a GPU before making code changes.
  • Advanced ray tracing – New features, including conical ranging, denoising of ancillary features, and support for FP16, provide more robust shapes and shorten render times, improving overall rendering performance. Support for upcoming Intel Xe GPUs along with real-time denoising further improves the industry-leading production quality rendering of the final image.
  • Extensive development environment support including deeper integration of Microsoft Visual Studio Code, support for Microsoft Visual Studio 2022, and Microsoft WSL2 for Linux development on Windows.

Comments are closed.