Microsoft, Meta and Intel announce AI initiative to rival Nvidia. What is it and why is it important?

Top tech companies like Meta, Microsoft, Advanced Micro Devices, Broadcom, Cisco Systems, Hewlett Packard Enterprise and Intel announced on Thursday that they have developed a new industry standard called “Ultra Accelerator Link,”. The new standard is an attempt to establish communication between artificial intelligence accelerators, systems that can help process vast amounts of data used in AI-related tasks.

Also Read | Apple and OpenAI forge billion-dollar AI partnership ahead of WWDC 2024: Report

Explaining the aim of new initiative, the companies said in a statement, "An industry specification becomes critical to standardize the interface for AI and Machine Learning, HPC (high-performance computing), and Cloud applications for the next generation of AI data centers and implementations,"

"By creating an interconnect based upon open standards, UALink will enable system OEMs, IT professionals and system integrators to create a pathway for easier integration, greater flexibility and scalability of their AI-connected data centers." the statement further stated

Why are top AI companies creating a new standard? 

The initiative is being seen as an attempt to challenge Nvidia's dominance in the AI chip market. Notably, the Jensen Huang led company holds around 80% of the market share in AI chip market which is a cause of  concern for top AI players like Microsoft, Meta and others.

Also Read | Explained: Why everyone’s talking about the Microsoft AI Copilot+ PCs

Notably, a report from Bloomberg earlier this year had also suggested that OpenAI CEO Sam Altman was talking to a number of potential investors in a bid to start a chip-making company that could run the company's AI products like ChatGPT, DALL-E and more. 

The report had stated that OpenAI had not yet finalised their move yet, and was discussing various options to solve the AI chip shortage. OpenAI was also said to be focusing on diversifying its suppliers beyond Nvidia and working with other chip manufacturers.