Happy Birthday Unified Acceleration Foundation (UXL)
One year ago, leading contributors across the hardware and software industry came together to foster an open multiarchitecture multivendor ecosystem. Intel is proud to be one of the founding steering committee members of the Unified Acceleration Foundation (UXL) along with ARM*, Fujitsu* Google Cloud*, Imagination Technologies*, Qualcomm Technologies Inc.*, Samsung*, and VMware* by Broadcom*.
Why was the UXL Foundation needed? With the growth of AI and accelerated computing in devices ranging from automobiles and edge devices to PCs to the largest data centers in the world, developers, in the ecosystem recognized a series of technical and economic challenges. Technically, proprietary programming models deliver code that is biased to a single architecture or product implementation The technical definition is closed and controlled by the product needs of a single vendor rather than an ecosystem. Economically, the costs of applications being locked to a single vendor or architecture limits choice and presents supply chain, cost, and related risks.
The open governance principles of the Linux Foundation’s Joint Development Foundation, addresses these challenges by incorporating and addressing the diverse perspectives and requirements of players in AI, Embedded & Edge Compute, as well as High Performance Computing through workgroups developing scalable and standardized programming languages, performance library interfaces, and specifications.
Development areas for a common unified parallel compute ecosystem, built around open standards, range from software programming framework and platform topics including SYCL*, OpenCL™ software, math, and AI all the way to integrating those approaches and models with hardware architectures from different vendors and safety-critical system needs.
The vision that is quickly becoming a reality is one of a fully scalable open source developer ecosystem for all aspects of accelerated computing.
The UXL Foundation is establishing the “triple crown” for accelerator software: open source, open standard, and open governance. Join us as we create a software ecosystem that unlocks the full potential of accelerators!
— Robert Cohn, Senior Principal Engineer, Intel Corporation
Foundational Elements
This approach’s foundational elements are close collaboration and alignment with the SYCL and OpenCL projects at Khronos Group* and UXL Foundation’s other affiliate partner, the Autoware Foundation*.
- SYCL provides the open source basis for cross-platform acceleration and abstraction.
- The Autoware Foundation applies the same principles to safety critical AI and autonomous driving.
The oneAPI software platform and open multiarchitecture programming model with its community of over 150 participants, libraries and specifications complete the rich set of building blocks the UXL Foundation is expanding on:
Figure 1: oneAPI Specification Elements
A Year of Rapid Growth and Evolution
Since its inception, membership in the UXL Foundation has grown rapidly, including over 30 members, silicon vendors, software vendors, original design manufacturers, AI solution providers, and automotive companies. The industry initiative is growing with increased use of the oneAPI programming model in Artificial Intelligence, Visual Computing, HPC, Edge, and other segments. Fujitsu and Arm leverage the UXL Foundation for AI initiatives, the Autoware Foundation and The UXL Foundation Collaborate on Achieving Software Portability for Open Source Autonomous Driving Stack.
Join us October 9th-10th for an in-depth showcase of the oneAPI specification and open-source projects that implements them. These projects are designed to enable developers to use a single code base across multiple accelerators and architectures to enable artificial intelligence, high performance computing, edge compute, automotive, and more.
Hosted by the UXL Foundation, this community led conference brings together developers to explore, share, and showcase the capabilities of oneAPI through hands-on tutorials, demos, technical talks, panel discussions and more!
Join us for two days across global time zones on how oneAPI is being used today and where it is going and join our ever-growing community.
Figure 2: UXL oneAPI Dev Summit Announcement
Open Parallel Developer Community
The UXL Foundation is not only a place where industry leaders and developers define and discuss open standards for heterogeneous accelerated compute throughout the modern world; a world where AI-infused solutions for everything from health, science, finance, safety, data analysis, autonomous driving, and everyday household convenience surround us.
It is also a rich and active software developer community of contributors to:
- oneAPI Elements Open Source Projects
- The SYCLomatic CUDA to SYCL Migration Tool
- oneAPI Code Samples
- Awesome List of Hundreds of oneAPI and SYCL Projects
- CUDA* to SYCL* Catalog of Ready-to-Use Applications
The oneAPI developer community, set of software development tools, and set of ready-to-deploy SCYL based solutions is a rich tapestry of contributions representative of the wide applicability of its open parallel standards-based approach across the software industry.
Intel’s Commitment to oneAPI and the UXL Foundation
Intel has been a champion of open source software development with contributions to many projects from GNU*, LLVM* and CLANG* to OpenCL™, OpenMP* and UEFI* for many decades. In that very spirit, we started the journey that led to the UXL Foundation with its embrace of the oneAPI open, cross-industry, standards-based, multi-vendor programming model in 2019, when we launched the oneAPI industry initiative.
We are glad to see its core principles and open source spec element contribution embraced by the wider industry and continue to thrive as one of its central building blocks.
oneAPI provides a comprehensive set of libraries, open source repositories, SYCL-based C++ language extensions, and optimized reference implementations to accelerate the following goals:
- Define a common, unified, open multiarchitecture and multivendor software platform.
- Ensure functional code and performance portability across hardware vendors and accelerator technologies.
- Enable an extensive set of specifications and library APIs to cover programming domain needs across industries and compute and AI use cases.
- Meet the needs of modern software applications that merge high-end computational needs and AI.
- Provide a developer community and open forum to drive a unified API for a unified industry-wide multiarchitecture software development platform.
- Encourage ecosystem collaboration on the oneAPI specification and compatible oneAPI implementations.
Celebrate with New oneAPI Centers of Excellence
The UXL Foundation’s rapid growth is not only fueled by our fellow members in the technology industry. Research institutions and universities worldwide have embraced the UXL Foundation and oneAPI to accelerate and scale their challenging workloads at the forefront of health, science, and data analytics. Many of their solutions leverage Generative AI to catalog vast knowledge libraries or Predictive AI for results interpretation and simulation. They are leading the effort of immersing computer engineering students and the next generation of technology leaders in the modern open software development paradigms for an AI-centric world of parallel offload acceleration.
The oneAPI Academic Centers of Excellence consist of over 30 research institutions and universities, ranging from Argonne National Labs to Zuse Institute Berlin and from the University of Bristol to the University of Utah. Many more individual researchers participate in the Intel® Innovators Program of oneAPI.
Two new members are joining the oneAPI Academic Centers of Excellence this month. They will leverage oneAPI for research and learning with AI.
LAION oneAPI Center of Excellence for Personalized AI Education
Large-scale Artificial Intelligence Open Network (LAION) is a non-profit organization that aims to make large-scale machine learning models, datasets, and related code available to the general public. It is becoming a oneAPI Center of Excellence to help it advance the development of BUD-E (Buddy for Understanding and Digital Empathy), an open-source, empathetic AI education assistant that aims to democratize personalized learning worldwide.
Leveraging open, standards-based oneAPI multiarchitecture programming, BUD-E is designed to provide personalized education to learners worldwide, with a special focus on children in developing countries. The project aims to create an AI companion that can interact with warmth, empathy, and high emotional intelligence, encouraging learners to explore the world and discover their potential.
We are thrilled to establish this new oneAPI AI Center of Excellence with Intel. As AI continues to reshape the educational landscape, there is a pressing need to ensure that these powerful tools are accessible to all, regardless of their economic or geographical circumstances. With oneAPI, we can optimize BUD-E’s performance across various hardware architectures, making high-quality AI-assisted education a reality for millions of children worldwide.
— Christoph Schuhmann, Founder, Organization Lead, LAION e.V.
University of Illinois Chicago oneAPI Center of Excellence for AI and HPC in Scientific Discovery
University of Illinois Chicago is becoming a oneAPI Center of Excellence to advance computational science education for AI and next-generation heterogeneous programming and empower students to optimize code for parallel architectures and high-performance multiarchitecture environments. Integrating cutting-edge tools and technologies, they thus intend to foster a versatile, technologically adept workforce capable of leveraging AI and HPC for scientific discovery.
The AI/oneAPI Centers of Excellence are helping the next cohort of developers to drive future software innovation using the power of CPUs, GPUs, and other accelerators. Our collaboration with Intel provides real-world experience and opportunities for our students.
— Dr. Michael E. Papka, the Warren McCulloch Professor of Computer Science at UIC and Director of the Argonne Leadership Computing Facility
Join the Future of Parallel Software Development
This is the best time to join the UXL Foundation’s movement towards open parallel compute and AI for all. AI leveraging accelerated compute is finding its way into every facet of technology development. By embracing openness, you can access the active ecosystem of software developers, the resources of a wider community, and the feedback of fellow travelers.
Let us drive the future of artificial intelligence and accelerated computing together as part of the UXL Foundation today:
Join us October 9th-10th:
Get oneAPI Development Tools
In addition to SYCL-enabled versions of the CLANG compiler and math, as well as deep learning and math library APIs, Intel provides a complete, ready-to-use binary distribution of software development tools geared toward multiarchitecture development targeting oneAPI and UXL Foundation projects.
Find out more about Intel’s approach towards supporting the UXL Foundation and oneAPI at our oneAPI Overview page.
Develop and scale your code with confidence. Get started today and download the Intel® oneAPI Base Toolkit.
Additional Resources
Joining UXL Foundation
oneAPI Ecosystem
- What is oneAPI? Demystifying oneAPI for Developers
- AI and oneAPI Academic Centers of Excellence
- Catalog of Ready-to-use SYCL Applications
- Awesome List of oneAPI Projects
Code Migration to SYCL