From The Hindu: [published_on]

Nvidia has reached a $2.2 trillion market cap by producing AI chips that power generative AI developers like Microsoft and OpenAI. A coalition including Qualcomm and Google plans to challenge Nvidia’s dominance with an open-source project called UXL, aiming to create software that can run on any AI accelerator chip.

The UXL Foundation, backed by tech companies like Intel and Google, aims to develop software that can power various AI accelerator chips and run on any machine. The open-source initiative seeks to challenge Nvidia’s dominance by offering developers more choice in hardware and promoting an open ecosystem for machine learning frameworks.

UXL plans to address computing problems dominated by a few chipmakers, focusing on AI applications and high-performance computing. The foundation aims to attract a critical mass of developers to its platform and eventually support Nvidia hardware and code. This initiative follows an influx of over $4 billion into 93 efforts aimed at unseating Nvidia in the AI software space.

Venture financiers and corporate dollars have invested over $4 billion in efforts to challenge Nvidia’s hold on AI software in the last year. Startups received over $2 billion in funding in 2023 to develop alternatives to Nvidia’s widely-used CUDA software, highlighting a growing interest in breaking Nvidia’s dominance in AI technology.



Read more at The Hindu: Behind the plot to break Nvidia’s grip on AI by targeting software