Joint Development foundation Project
UXL

Unified Acceleration Foundation

  • Build a multi-architecture multi-vendor software ecosystem for all accelerators.
  • Unify the heterogeneous compute ecosystem around open standards.
  • Build on and expand open-source projects for accelerated computing.
Join Participate

UXL Foundation Webinar

Join us on April 30th at 8:00 am PDT, for a Complimentary Live Webinar presented by the UXL Foundation. Learn about the UXL Foundation and meet members of our Steering Committee to learn more about oneAPI and why these companies have come together to use open source elements of oneAPI to drive solutions for their businesses.

This webinar will cover:

  • What are the UXL Foundation focus areas?
  • What is the oneAPI Specification?
  • What target markets are specific companies targeting for solutions based on oneAPI?
  • What open source project momentum is there and where can we learn more?
Register Now

Steering Members

Imagination Tech

Steering Committee Members

Andrew Wafaa

Senior Director Software Communities & Fellow

“”

Rod Burns

VP of Ecosystem

Masahiro Doteguchi

Senior Engineering Manager

Fujitsu

Dr. Priyanka Sharma

Director of Software Engineering

Fujitsu

Penporn Koanantakool

Senior Software Engineer

Dave Murray

AI Software Product Management

Imagination Tech
“”

Robert Cohn

Senior Principal Engineer

Intel
“”

Dr. Vinesh Sukumar

Head of AI/ML Product Management

Qualcomm

Hanwoong Jung

Project Leader of Universal Deep Learning Compiler (UDLC)

Samsung

Ramesh Radhakrishnan

Technical Director , OCTO HPC/ML

VMWare

Spec Elements

The Spec is made of 7 core elements.

oneDPL

oneAPI Data Parallel C++ Library

A companion to the DPC++ Compiler for programming oneAPI devices with APIs from C++ standard library, Parallel STL, and extensions.

oneDNN

oneAPI Deep Neural Network Library

High performance implementations of primitives for deep learning frameworks.

oneCCL

oneAPI Collective Communications Library

Communication primitives for scaling deep learning frameworks across multiple devices.

Level Zero

oneAPI Level Zero

System interface for oneAPI languages and libraries.

oneDAL

oneAPI Data Analytics Library

Algorithms for accelerated data science.

oneTBB

oneAPI Threading Building Blocks

Library for adding thread-based parallelism to complex applications on multiprocessors.

oneMKL

oneAPI Math Kernel Library

High performance math routines for science, engineering, and financial applications.

UXL News

Man on computer, city in the background

UXL Foundation readying alternative to Nvidia's CUDA for this year

The UXL Foundation is readying its open standard accelerator programming model, touted by some as an alternative to Nvidia’s CUDA platform, for “a spec release in Q4.

Learn More
Article

Vmware Collaborates With Intel To Unlock Private AI Everywhere

Article

Unified Acceleration (UXL) Foundation

Article

The Unified Acceleration Foundation Aims to Boost ML, AI Workloads No Matter the Architecture

Article

UXL Foundation Safety Critical SIG Created

Hear more from the Steering Members

“The foundation members aspire to build the largest open ecosystem for accelerated computing. The initial contributions to the foundation will bring an existing open standards based platform with open governance. Our ultimate aim is to foster a multi-architecture and multi-vendor programming platform for all accelerators”, said Rod Burns, VP Ecosystem, Codeplay Software and UXL Foundation Steering Committee Chair.

Read More from Rod Burns, VP Ecosystem, Codeplay Software and UXL Foundation Steering Committee Chair

“The future of human centric, sustainable digital transformation needs a holistic approach to technology development and open collaboration with the international community. The Unified Acceleration Foundation’s innovative strategy of building a unified development ecosystem will play a vital role in advancing the creation of cross-platform software and accelerators”, said Hirai Kouichi, Head of Software Development Division under the Advanced Technology Development Unit, Fujitsu.

“The growing diversity of accelerators presents a tremendous opportunity for the industry but can also introduce challenges. Consistent programming models help developers write code that can run efficiently on any accelerator, fostering innovation and choice,” said George Elissaios, Sr. Director, HPC and ML infrastructure at Google Cloud.

“The technology that is improving our day-to-day experiences of driving, entertainment, healthcare and more, is increasingly data-intensive and complex. A new open and collaborative approach to computing is required to provide the necessary acceleration in an efficient and performant manner, whether in the cloud or at the edge,” said Shreyas Derashri, Vice President of Compute Product Management at Imagination Technologies. “As a founding member of the Unified Acceleration Foundation, Imagination Technologies will help unite the technology ecosystem around the oneAPI spec and encourage its widespread adoption for software acceleration.”

“As a new independent entity, the Unified Acceleration Foundation will help build an open source programming model across a wide range of architectures and vendors,” said Vinesh Sukumar, Senior Director of Product Management, Qualcomm Technologies, Inc. “This helps to address a critical need in the industry for flexible and open source alternatives that will enable a broad range of unique needs.”

“Samsung is pleased to be part of the Unified Acceleration Foundation, which will help expand the basis and adoption for memory coupled computing technologies, such as PIM and PNM”, said Seungwon Lee, Master and lead of computing software team at Samsung Advanced Institute of technology(SAIT). “We expect the open collaboration will also bolster programming models and software stacks to effectively accelerate AI/HPC applications in future datacenter and supercomputing systems.”

“Linux and GNU transformed the CPU software stack through open source and standards, encompassing everything from embedded to cloud computing. As a founding member of the Unified Acceleration Foundation, I believe that open source and standards are essential for creating a cross platform software stack for GPU’s and other accelerators that will serve as the foundation for the next generation of computationally and data-intensive applications”, said Robert Cohn, Intel Corporation, oneAPI Specification Editor.

The Unified Acceleration Foundation is an evolution of the oneAPI initiative and focuses on the development of a specification and open source projects through Working Groups and Special Interest Groups (SIGs).

Specification

The specification includes 7 core elements of oneAPI, developed by a core team with feedback from the community.

Open Source

oneAPI allows developers to make accelerator choices.

Community

We invite radical collaboration from across the ecosystem to create a shared industry spec.

And much more...