In order to break NVIDIA's dominance in the AI field, an industry group including Intel, Google, Fujitsu, Arm, etc. is building an AI development environment to compete with CUDA

NVIDIA has a large share in GPUs used for AI development, and its software development and execution environment ``CUDA'' is also widely adopted in AI development sites. In order to counter NVIDIA's single-win situation, the industry organization Unified Acceleration Foundation (UXL Foundation) , which includes technology companies such as Intel, Fujitsu, Google, and Arm, is building an open software development environment.

UXL Foundation: Unified Acceleration
https://uxlfoundation.org/

Unified Acceleration (UXL) Foundation
https://www.intel.com/content/www/us/en/developer/articles/news/unified-acceleration-uxl-foundation.html

Exclusive: Behind the plot to break Nvidia's grip on AI by targeting software | Reuters
https://www.reuters.com/technology/behind-plot-break-nvidias-grip-ai-by-targeting-software-2024-03-25/

NVIDIA's GPUs are widely used in AI development sites due to their high processing performance, and NVIDIA's software development and execution environment 'CUDA' is also used in many development sites. Additionally, since CUDA is a system that only targets NVIDIA GPUs, there are situations in which developers who are familiar with CUDA development are forced to choose NVIDIA GPUs.

The UXL Foundation is an industry organization that aims to build a new software development environment based on the oneAPI software development environment led by Intel. It was founded by companies such as Qualcomm, Samsung, and VMware. In addition, the UXL Foundation was established as a subsidiary organization of the Linux Foundation, and has declared that it will keep its technical specifications and software source code open.

While CUDA only targets NVIDIA GPUs, oneAPI does not limit compatible manufacturers and supports a wide variety of hardware such as CPUs, GPUs, FPGAs, and AI accelerators. Also, while CUDA is a closed source system, oneAPI is developed as an open source project. By further developing oneAPI, the UXL Foundation aims to create an environment where there is no need to consider the type of processor targeted when developing software.

UXL is an evolution of the #oneAPI initiative, which has paved the way for UXL as an open programming model with #CPU , #GPU , FGPA, and accelerators. UXL is all this and more.

They've all come together to drive open accelerated compute and cross-platform performance.

— UXLfoundation (@UXLfoundation) March 13, 2024

Until now, the UXL Foundation had not specified when the specifications for the development environment would be finalized, but on March 26, 2024, Reuters reported that ``The UXL Foundation is preparing to finalize the specifications in the first half of 2024.'' . Furthermore, it has been reported that the UXL Foundation has already received technical contributions from third parties and that there are plans to support NVIDIA hardware. Vinesh Sukumar, head of AI and machine learning at Qualcomm, told Reuters: ``We're showing developers how to migrate away from NVIDIA's platform.'' It shows a stance against CUDA.