The oneAPI initiative, governed by the UXL Foundation*, turns the vision of a unified, standards-based, open programming model for accelerated computing into reality. oneAPI provides freedom from being locked into a single vendor’s proprietary architecture and toolchain through a cross-architecture, cross-vendor CUDA* alternative to target diverse hardware including CPUs, GPUs, and other accelerators not just from Intel but also NVIDIA*, AMD*, ARM*, and others. It does so while offering a rich portfolio of libraries, tools, compilers, and ecosystem software integrations to maximize both developer productivity and application performance.
Joe Curley, vice president and general manager of Software Products and Ecosystem at Intel, provides a concise and compelling overview of the oneAPI vision and its benefits to the software developer community targeting high-performance, compute-intensive, multiarchitecture execution environments for HPC, AI, and rendering applications. He details how an open programming model that delivers both accelerated application performance and enhanced developer productivity in a multiarchitecture and multivendor solution prevents lock-in and allows for easy code reuse and portability across diverse current and future architectures.
The oneAPI initiative is evolving into the Unified Acceleration (UXL) Foundation, formed under the Linux Foundation umbrella, further embracing the power of open collaboration. Intel in partnership with industry leaders such as Arm, Fujitsu*, Google*, Imagination*, Qualcomm*, Samsung*, and others intends to unify the heterogeneous compute ecosystem around open standards.