The UXL team will develop a suite of software to challenge Nvidia's CUDA

The Unified Acceleration Foundation (UXL) will develop a competitor for Nvidia’s CUDA. Their main goal is to offer developers an open-source suite of software to make them switch sides. However, the goal is hard to attain because four million developers use the Nvidia CUDA architecture. After all, the company’s chips are superior to the ones made by other chipmakers. Yet, they are hard to come by due to their high demand.

Who is the UXL group?

The UXL group is a team of tech companies, and some of them are Intel, Google, and Samsung. Together, they work on a big project to develop a software suite similar to Nvidia’s CUDA but far less restraining for developers.

For this goal, Intel shared its OneAPI for the project, which removes barriers to hardware access. Thus, the new software will allow developers to run their code on various devices and chips. Also, with it, they will no longer need specific coding languages and code bases. In addition, the project will develop new tools for them. So, there is no longer any tie between developers and Nvidia’s tech.

However, the UXL group will support Nvidia when the time comes and try to partner with the company to create better chips. Thus, their other goal is to improve the whole chipmaking industry to power up future AI hardware. On top of that, the team seeks to aid other chipmakers and cloud-computing companies to ensure the deployment of their software and tools. Also, while Microsoft and AMD aren’t a part of the UXL group, they are trying to challenge Nvidia’s monopoly by working together on AI chips.

According to Reuters, Nvidia is looking forward to having competitors. After all, if developers gain access to new technologies, they will improve the industry quickly. Thus, Nvidia and other companies will discover more innovations in the field.

The UXL group plans to identify the technical specifications needed for the project during the first half of this year and make the software functional for developers by the end of the year.

In summary, the UXL group aims to bring a new open-source software suite to challenge Nvidia’s supremacy over AI chips. The UXL software will allow developers to use their code on different devices and chips, thus no longer needing Nvidia’s CUDA platform. Startups, small companies, and others will also have more chances to join the AI race.

What are your thoughts? Are you going to try the UXL software? Let us know in the comments.