On the Open Supply Summit Europe in Bilbao, Spain, the Linux Basis at this time introduced the launch of the Unified Acceleration (UXL) Basis. The group’s mission is to ship “an open commonplace accelerator programming mannequin that simplifies growth of performant, cross-platform functions.”
The muse’s founding members embrace the likes of Arm, Fujitsu, Google Cloud, Creativeness Applied sciences, Intel, Qualcomm and Samsung. The corporate most conspicuously lacking from this record is Nvidia, which gives its personal CUDA programming mannequin for working with its GPUs.
At its core, this new basis is an evolution of the oneAPI initiative, which can also be aimed to create a brand new programming mannequin to make it simpler for builders to assist a variety of accelerators, irrespective of whether or not they’re GPUs, FPGAs or different specialised accelerators. Like with the oneAPI spec, the purpose of the brand new basis is to make sure that builders could make use of those applied sciences with out having to delve deep into the specifics of the underlying accelerators and the infrastructure they run on.
OneAPI was backed by Intel and others and this new basis now plans to take the oneAPI specification and assist implement it throughout the trade. That’s what the Linux Basis does finest, in spite of everything, with a give attention to its open governance mannequin that attracts a line between the enterprise and technical decision-making processes.
“Linux and GNU reworked the CPU software program stack by open supply and requirements, encompassing all the things from embedded to cloud computing,” mentioned Robert Cohn, Intel Company, oneAPI specification editor. “As a founding member of the Unified Acceleration Basis, I imagine that open supply and requirements are important for making a cross-platform software program stack for GPUs and different accelerators that may function the basis for the following era of computationally and data-intensive functions.”