Intel® oneAPI DPC++/C++ Compiler
Talk to fellow users of Intel® oneAPI DPC++/C++ Compiler and companion tools like Intel® oneAPI DPC++ Library, Intel® DPC++ Compatibility Tool, and Intel® Distribution for GDB*

oneAPI 2023.1 release is now AVAILABLE!

Barbara_P_Intel
Moderator
707 Views

oneAPI 2023.1 release is AVAILABLE NOW!


Intel® oneAPI Base Toolkit

  • Intel® oneAPI DPC++/C++ Compiler adds auto cpu_dispatch, kernel properties for SYCL, more SYCL* 2020 features to improve developer productivity and boost CPU and GPU offload performance. Bfloat16 is now a full feature (out of the experimental namespace) in the compiler. More OpenMP* 5.0 and 5.1 compliance, new standard features, and improves performance. The compiler enhances developer efficiency with better compiler error messaging and error handling for SYCL and OpenMP code.
  • For easier CUDA-to-SYCL code migration, Intel® DPC++ Compatibility Tool (based on the open source SYCLomatic project) adds support for CUDA 12.0 headers, and migration of more CUDA APIs to the equivalent SYCL language and oneAPI library functions including runtime, math, and neural network domains.
  • Identify performance gained from High Bandwidth Memory on Intel Data Run Intel® VTune Profiler for each mode (HBM only, Flat, Cache) to identify which profile offers the best performance.
  • Use Intel® VTune Profiler for visibility into XeLink cross-card traffic on Intel® Data Center GPU Max Series for issues such as stack-to-stack traffic, throughput and bandwidth bottlenecks. Identify imbalances of traffic between CPU and GPU through a GPU topology diagram.
  • Intel® oneAPI Math Kernel Library improves Intel® Data Center GPU Max Series performance via new real FFTs plus 1D and 2D optimizations, new random number generators, as well as additional Sparse BLAS and LAPACK inverse optimizations for Cholesky, triangle matrix, and batch LU routines.
  • The Intel® oneAPI DPC++ Library has improved performance of the sort, scan, and reduce algorithms.
  • The Intel Distribution of GDB* adds debug support for Intel® Arc™ Graphics on Windows* via the Level 0 debug API, and improves the debug performance on Linux* for discrete GPUs.

Intel® oneAPI HPC Toolkit

  • Intel® Fortran Compiler enhanced OpenMP 5.0, 5.1 compliance, and improved performance.
  • Intel® MPI Library improves performance for collectives using GPU buffers and through default process pinning on CPUs with E-cores and P-cores
  • Intel® Cluster Checker is now deprecated and will be discontinued in a future release.

Intel® oneAPI Rendering Toolkit

  • Intel® Embree v4.0 brings common code base Beta ray tracing support for Intel® Arc™ GPUs while maintaining GOLD support for CPU platforms.
  • Increased sampling quality and, rendering efficiency is achieved by implementing state-of-the-art path guiding methods included in the newest library addition, Intel® Open PGL.
  • Updates to Intel® OSPRay Studio extend the glTF “container” to include both geometry and volume in a single scene and enhancements to the wavelet slice generator allows much larger dataset generation.


Download Intel® oneAPI Toolkits TODAY!

0 Kudos
0 Replies
Reply