YOU ARE AT:AI InfrastructureOpenAI set to start mass production of its own AI chips

OpenAI set to start mass production of its own AI chips

OpenAI will start production next year with U.S. semiconductor company Broadcom

In sum – what to know:

OpenAI partners with Broadcom for chips – The company plans mass production of custom AI processors in 2026 to reduce reliance on Nvidia and secure capacity for its own workloads.

Broadcom secures $10bn deal – CEO Hock Tan confirmed a mystery new customer order, widely reported as OpenAI, which pushed Broadcom shares up 15% and its valuation to $1.7 trillion.

Industry trend toward in-house chips – Following Google, Amazon, and Meta, OpenAI joins peers in developing dedicated hardware to meet rising AI compute demands and better manage performance, costs, and supply risks.

U.S. AI company OpenAI is getting ready to begin large-scale production of its own artificial intelligence chips next year, in partnership with compatriot semiconductor company Broadcom, the Financial Times reported, citing sources familiar with the plan.

The report noted that the initiative is aimed at easing OpenAI’s reliance on U.S. chipmaker Nvidia hardware while meeting surging demand for computing capacity to train and run AI models. The chips are expected to be used internally rather than sold to external customers, according to the report.

Broadcom’s CEO Hock Tan referenced a new unnamed client committing to $10 billion in chip orders, without disclosing the company’s identity. Shares in Broadcom rose 15% in early trading Friday, lifting its market value to about $1.7 trillion. People close to the matter confirmed OpenAI as the customer, although both companies declined to comment, the report stated.

The move follows similar strategies from other major technology giants — including Google, Amazon, and Meta — that have developed their own specialised processors for AI workloads. OpenAI initially began working with Broadcom last year, but details of a production timeline had remained uncertain until now.

Tan told analysts that Broadcom had secured a fourth major client for its custom AI chip business, which he said had created “immediate and fairly substantial demand.” He added that shipments for this customer would begin strongly from next year.

In the United States, OpenAI is working with SoftBank and Oracle on its Stargate initiative, which aims to deliver 4.5 gigawatts of capacity — an initiative unveiled by President Donald Trump in January.

The Stargate initiative has an estimated long-term cost of $500 billion; $100 billion has already been earmarked for its early phases. At its core, the Stargate scheme plans to build a network ofAI data centers across the U.S. The initiative aims to provide sufficient capacity to meet growing demand for AI across sectors including scientific research, healthcare, automation, defense, and finance

The project could become the largest AI infrastructure buildout to date, spanning multiple locations and phases. It is being led by OpenAI in partnership with SoftBank and Oracle. OpenAI is contributing AI models and technical leadership, while SoftBank is providing funding and strategic support. Oracle is a key infrastructure partner.

OpenAI previously announced the launch of “OpenAI for Countries” as an initiative within the Stargate project. It has announced plans to offer formalized infrastructure partnerships with national governments, in coordination with the U.S. administration, to help build local AI data center capacity.

Under this initiative, the company will partner with countries to help build in-country data center capacity. These secure data centers will help support the sovereignty of a country’s data, build new local industries, and make it easy to customize AI and leverage their data in a private and compliant way, OpenAI said.

OpenAI is planning to construct a massive new data center in India as part of its global Stargate AI infrastructure expansion, according to recent press reports.

The reports, which cited people familiar with the matter, noted that the ChatGPT developer is in discussions with potential local partners to establish a facility with at least 1 gigawatt of capacity in the world’s most populous nation.

If realized, the site would rank among India’s largest, alongside recent data center investments by Microsoft, Google, and Reliance Industries.

ABOUT AUTHOR

Juan Pedro Tomás
Juan Pedro Tomás
Juan Pedro covers Global Carriers and Global Enterprise IoT. Prior to RCR, Juan Pedro worked for Business News Americas, covering telecoms and IT news in the Latin American markets. He also worked for Telecompaper as their Regional Editor for Latin America and Asia/Pacific. Juan Pedro has also contributed to Latin Trade magazine as the publication's correspondent in Argentina and with political risk consultancy firm Exclusive Analysis, writing reports and providing political and economic information from certain Latin American markets. He has a degree in International Relations and a master in Journalism and is married with two kids.