Today’s workloads benefit from specialized hardware architectures. However, these architectures have typically required unique languages and tools, which limits code reuse — narrowing hardware choices and hindering adoption of innovative architectures. oneAPI is an open, unified programming model built on standards to simplify development and deployment of data-centric workloads across CPUs, GPUs, FPGAs and other accelerators.
On February 16 Intel software development and public sector experts shared how the oneAPI industry initiative delivers on bringing an open, unified cross-architecture programming to the ecosystem — providing an alternative to proprietary programming models. They also discussed how Intel’s oneAPI toolkits, along with the Intel® DevCloud, provide the production tools needed to accelerate advances into distributed intelligence era and extract the most application performance on multiple types of Intel® architecture.
