https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/Intel-oneAPI-DPC-C-Compiler-2026-1-is-available/m-p/1753432#M4771 <P>Intel® oneAPI DPC++/C++ Compiler 2026.1 has been released and is available for download <A href="https://www.intel.com/content/www/us/en/developer/tools/oneapi/dpc-compiler-download.html" target="_self">here.</A></P> <P>New Features and Enhancements in this release:</P> <UL> <LI><STRONG>Expanded SYCL capabilities and performance:</STRONG> Added new host_task overloads to&nbsp;sycl_ext_oneapi_enqueue_functions, introduced native recording mode for sycl_ext_oneapi_graph, extended Intel Xe GPU device queries, and significantly improved get_kernel_info performance.</LI> <LI><STRONG>Enhanced OpenMP and GPU offload support:</STRONG> Added -fopenmp-loop-threadization-level for loop performance tuning, introduced support for requires unified_shared_memory to simplify CPU-to GPU ports, and improved offload performance for workloads using target teams if(cond).</LI> <LI><STRONG>Compiler and language runtime improvements:</STRONG> Improved handling of volatile variables with setjmp, and updated device compilation behavior so SYCL/OpenMP offload no longer forces preserve-sign denormal flushing by default while maintaining host compatibility and user overrides.</LI> <LI><STRONG>New architecture optimizations:</STRONG> Implemented Intel® AVX10.2 optimizations in SVML, delivering substantial performance gains for vectorized mathematical functions on AVX10.2-enabled platforms.</LI> </UL> Wed, 08 Jul 2026 20:44:41 GMT Sravani_K_Intel 2026-07-08T20:44:41Z https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/Intel-oneAPI-DPC-C-Compiler-2026-1-is-available/m-p/1753432#M4771 <P>Intel® oneAPI DPC++/C++ Compiler 2026.1 has been released and is available for download <A href="https://www.intel.com/content/www/us/en/developer/tools/oneapi/dpc-compiler-download.html" target="_self">here.</A></P> <P>New Features and Enhancements in this release:</P> <UL> <LI><STRONG>Expanded SYCL capabilities and performance:</STRONG> Added new host_task overloads to&nbsp;sycl_ext_oneapi_enqueue_functions, introduced native recording mode for sycl_ext_oneapi_graph, extended Intel Xe GPU device queries, and significantly improved get_kernel_info performance.</LI> <LI><STRONG>Enhanced OpenMP and GPU offload support:</STRONG> Added -fopenmp-loop-threadization-level for loop performance tuning, introduced support for requires unified_shared_memory to simplify CPU-to GPU ports, and improved offload performance for workloads using target teams if(cond).</LI> <LI><STRONG>Compiler and language runtime improvements:</STRONG> Improved handling of volatile variables with setjmp, and updated device compilation behavior so SYCL/OpenMP offload no longer forces preserve-sign denormal flushing by default while maintaining host compatibility and user overrides.</LI> <LI><STRONG>New architecture optimizations:</STRONG> Implemented Intel® AVX10.2 optimizations in SVML, delivering substantial performance gains for vectorized mathematical functions on AVX10.2-enabled platforms.</LI> </UL> Wed, 08 Jul 2026 20:44:41 GMT https://community.intel.com/t5/Intel-oneAPI-DPC-C-Compiler/Intel-oneAPI-DPC-C-Compiler-2026-1-is-available/m-p/1753432#M4771 Sravani_K_Intel 2026-07-08T20:44:41Z