Intel announced the launch of the oneAPI toolkit 2022. The new version of the toolkit expands the features of different architectures, providing developers with more utilities and architecture options for accelerating computing. Intel said that version 2022 of oneAPI is ready for multi-language support and heterogeneous architecture performance acceleration to further improve developer productivity on Intel platforms.
oneAPI is a cross-architecture programming tool designed to simplify programming across GPUs, CPUs, FPGAs, and AI accelerators. It can be used with Intel’s own devices or chips from other manufacturers to optimize workloads. oneAPI is a cross-platform programming interface. Customers can optimize and accelerate the processing program most efficiently for which specific architecture. It also supports Intel-specific instructions such as AVX-512 and DL Boost. Intel released the oneAPI toolkit at the end of last year.
Intel oneAPI toolkit 2022 provides many new features, including a unified compiler that implements C++, SYCL, and Fortran, CPU and GPU data-parallel programming that supports Python, accelerator performance modeling, and tuning, and performance acceleration for AI and ray-tracing visualization workloads. At the same, it supports advanced features of the latest and upcoming new hardware, including Intel’s 12th-generation Core series processors using AVX-VNNI, next-generation Sapphire Rapids Xeon Scalable processors, and various upcoming client and data center GPUs.
Currently, oneAPI can be
downloaded for free or used in Intel DevCloud.