Cerebras Files $3.5 Billion IPO Amid OpenAI Compute Deal

Avatar photo

ByLisa Grant

May 4, 2026

AI chipmaker Cerebras Systems seeks a $26.6 billion valuation while securing a massive compute partnership with OpenAI, as the industry pivots toward specialized infrastructure and away from Nvidia’s singular dominance.

The digital arms race has entered a high-stakes infrastructure phase as Cerebras Systems, a prominent challenger to Nvidia’s hardware hegemony, filed updated paperwork for a Nasdaq initial public offering. The company plans to sell 28 million shares priced between $115 and $125, aiming to raise up to $3.5 billion. This move would value the wafer-scale chip specialist at approximately $26.6 billion, a significant jump from its $23 billion private valuation just months ago.

Cerebras has successfully transitioned from a pure hardware vendor to a cloud service provider, a strategic pivot validated by a massive multi-year agreement with OpenAI. The deal, valued at over $20 billion through 2028, provides OpenAI with up to 750 megawatts of compute capacity. Financial disclosures reveal the efficacy of this shift, with Cerebras reporting fourth-quarter revenue of $510 million—a 76% year-over-year increase—and a net income of $87.9 million.

While Cerebras prepares for public markets, the competitive landscape for frontier models is facing new scrutiny. In recent court testimony, Elon Musk acknowledged that his AI venture, xAI, utilized distillation from OpenAI models to train its own systems. This practice of using a superior model’s output to refine a smaller one highlights the incestuous nature of current AI development and raises critical questions regarding intellectual property and the sovereignty of proprietary data sets.

State power is also consolidating around these commercial entities. The Pentagon recently finalized agreements with seven tech giants, including OpenAI, Google, Microsoft, and Nvidia, to deploy AI tools within classified military networks. Notably, xAI secured its seat at the table via SpaceX, while Anthropic was reportedly excluded following disagreements over usage restrictions. This fusion of private algorithmic power and national security infrastructure signals a deepening of the ‘Algorithmic State,’ where corporate black boxes become the backbone of defense.

Institutional shifts are further evidenced by the financial sector’s embrace of these technologies. CGI recently achieved Microsoft Copilot specialization to drive enterprise AI integration, while AI Interfaces, Inc. launched its KongXLM prediction engine. However, the rapid expansion of these systems has prompted warnings from U.S. Treasury Secretary Scott Bessent and international cybersecurity agencies, who caution that agentic AI and autonomous systems now pose a direct threat to financial stability and critical infrastructure.

As Cerebras prepares to test investor appetite for ‘boring’ infrastructure, the broader market is beginning to favor companies with tangible monetization and infrastructure leverage. This maturation of the AI trade suggests that while the initial hype of chatbots may be cooling, the battle for the physical and logical layers of the digital frontier is only beginning.

Leave a Reply

Your email address will not be published. Required fields are marked *