Meta expands Broadcom partnership to scale custom AI chip development

0
45
Meta and Broadcom deepen AI chip alliance with multi-gigawatt expansion plan
Meta and Broadcom deepen AI chip alliance with multi-gigawatt expansion plan

In a major push to strengthen its artificial intelligence infrastructure, Meta has expanded its collaboration with chip designer Broadcom to develop multiple generations of custom AI processors. The move comes as the company accelerates efforts to boost computing power across its platforms and support next-generation AI features.

Under the revised agreement announced on Tuesday, the partnership has been extended through 2029 and includes an initial commitment exceeding 1 gigawatt of computing capacity. This scale is estimated to be sufficient to power around 750,000 average US homes.

As part of the updated structure, Broadcom CEO Hock Tan will step down from Meta’s board and transition into an advisory role focused on the company’s custom chip strategy, according to a joint statement.

The expansion reflects a wider industry trend, where major tech firms including Meta, Google, and Amazon are increasingly designing proprietary chips to reduce dependence on Nvidia’s high-cost AI processors. The surge in demand for AI computing has significantly boosted Broadcom’s position in the semiconductor ecosystem, particularly through its work on custom processors and infrastructure software solutions.

Following the announcement, Broadcom shares rose 3.5% in extended trading, while Meta’s stock remained largely unchanged.

Meta CEO Mark Zuckerberg said the partnership supports the company’s long-term AI goals, stating it will help “build out the massive computing foundation we need to deliver personal superintelligence to billions of people.”

The company also noted that the deal represents the first phase of a broader, multi-gigawatt deployment plan. Broadcom’s Ethernet networking technology will be used to connect Meta’s expanding clusters of AI computing systems.

Meta’s chip roadmap includes its Meta Training and Inference Accelerator (MTIA) programme. The MTIA 300 already supports ranking and recommendation systems, while three additional chips are expected by 2027. Later versions will focus on AI inference, enabling models to process and respond to user queries more efficiently.

In a separate update, Meta also confirmed that board member Tracey Travis, who joined in 2020, will not seek re-election at the upcoming annual shareholder meeting.

Also read: Viksit Workforce for a Viksit Bharat

Do Follow: The Mainstream LinkedIn | The Mainstream Facebook | The Mainstream Youtube | The Mainstream Twitter

About us:

The Mainstream is a premier platform delivering the latest updates and informed perspectives across the technology business and cyber landscape. Built on research-driven, thought leadership and original intellectual property, The Mainstream also curates summits & conferences that convene decision makers to explore how technology reshapes industries and leadership. With a growing presence in India and globally across the Middle East, Africa, ASEAN, the USA, the UK and Australia, The Mainstream carries a vision to bring the latest happenings and insights to 8.2 billion people and to place technology at the centre of conversation for leaders navigating the future.