Intel oneAPI Initiative Evolves Into The Unified Acceleration "UXL" Foundation

Here's a surprise announcement I was briefed on last week and now made public by the Linux Foundation and Intel... The Linux Foundation is forming the Unified Acceleration (UXL) Foundation that is an evolution of Intel's oneAPI initiative and has the potential to make the compute accelerator ecosystem as a whole more open and unified across vendors.

The Unified Acceleration Foundation (UXL Foundation) is intended to be a cross-industry group to delivering an open standard programming model that simplifies the development of performant and cross-platform applications. This open standards for compute accelerators is bolstered by Intel's oneAPI and intended to garner industry adoption of the oneAPI specification.

Founding members of the UXL Foundation include Arm, Fujitsu, Google Cloud, Imagination Tech, Intel, Qualcomm Technologies, and Samsung. Notably absent from the UXL Foundation is AMD and to a lesser surprise NVIDIA. AMD for their part continues focused on the ROCm compute stack and NVIDIA obviously with CUDA. But via Intel's Codeplay Software acquisition work they've brought SYCL and various oneAPI elements already atop AMD and NVIDIA GPUs/drivers. It's also perhaps a bit surprising The Khronos Group wasn't involved in this effort given their industry standards expertise and ecosystem. In any event it will be interesting to see how the UXL Foundation evolves under the Linux Foundation.

Intel has been very open with oneAPI from the start and via their own contributions and the since-acquired Codeplay Software have been bringing various oneAPI toolkit components to work on non-x86 CPUs like ARM / POWER / RISC-V, various levels of AMD and NVIDIA GPU support for their programming model, and other efforts to ensure the oneAPI specification and Intel-developed software works well throughout the ecosystem. Intel's oneAPI efforts via open-source code and open standards around SYCL and friends has been an excellent effort the past several years and has been a pleasure watching it evolve.

Of the foundation, today's press release adds, "enabling solutions that are productive, performant, and provide customers with the freedom to choose the hardware that best fits their unique needs." The UXL Foundation will work on oneAPI / open accelerated compute across CPUs, GPUs, FPGAs, and other accelerators.

The Unified Acceleration Foundation website will be UXLFoundation.org (not yet live). This is quite an exciting development for an evolution of Intel's oneAPI but will be interesting to see how it ultimately plays out... There are some good names involved already but without the backing of AMD and NVIDIA, the GPU compute space at least will likely remain a fragmented mess. As mentioned there's been efforts anyhow to get Intel oneAPI software up and running on NVIDIA/AMD GPUs but the performance and level of support may be a different story -- especially when considering the limited hardware scope of AMD's ROCm stack and the like that run underneath. In any event I applaud today's announcement and I will certainly be monitoring the UXL Foundation progress on Phoronix.