Nvidia Corp.’s Chief Executive Officer, Jensen Huang, has unveiled plans to enable customers to use rival chips in data centers built around its technology. This strategic move recognizes the increasing trend of major clients like Microsoft Corp. and Amazon.com Inc. developing their own semiconductor capabilities.
Huang made the announcement at Computex in Taiwan, Asia’s largest electronics forum, during a nearly two-hour presentation that also highlighted the work of local supply chain partners. The key revelation was Nvidia’s new NVLink Fusion system, which facilitates the creation of more customized artificial intelligence infrastructure. This system combines Nvidia’s high-speed links with semiconductors from other providers for the first time, offering greater flexibility to customers.
The development marks a significant shift in Nvidia’s approach, as the company opens its ecosystem to other chipmakers. This move is likely to have far-reaching implications for the semiconductor industry, particularly for major players who are increasingly developing their own in-house semiconductor capabilities.
By allowing the integration of rival chips, Nvidia is catering to the evolving needs of its clients while maintaining its position as a leader in the AI infrastructure market. The NVLink Fusion system is set to provide data center operators with more options for building customized AI infrastructure, potentially accelerating the adoption of AI technologies across various industries.
