In a significant development within the artificial intelligence hardware landscape, OpenAI has secured an agreement with Cerebras Systems, an emerging competitor to Nvidia and Advanced Micro Devices, to provide an extraordinary 750 megawatts of computing power over the course of the next four years, concluding in 2028. This long-term collaboration enhances the operational deployment of OpenAI's gpt-oss open-weight models by leveraging Cerebras' specialized hardware platforms alongside established processors from Nvidia and AMD.
While the confidentiality surrounding the financial terms of this arrangement remains intact, industry sources speculate that the contract could be valued in excess of $10 billion. Such a substantial deal offers Cerebras a critical opportunity to broaden its revenue streams, which, until recently, were heavily concentrated with a single customer. Notably, in the first half of 2024, Cerebras' income was predominantly sourced from UAE-based conglomerate G42, accounting for approximately 87% of the company's total revenue.
The expansion of Cerebras' clientele forms a deliberate growth trajectory as articulated by the company's co-founder and CEO, Andrew Feldman. Feldman expressed in discussions that cultivating a small number of very large customers begins with satisfying an initial key client and subsequently expanding the customer base through maintaining strong partnerships. This methodical approach aims to solidify Cerebras’ presence in the competitive AI hardware market segment.
In parallel to this supply agreement, Cerebras' financial strategy has experienced notable shifts. The company previously announced plans for an initial public offering (IPO) but postponed these intentions in October, despite having reached a private valuation close to $8 billion. No specific rationale was provided for the decision to halt the IPO at that time, with the company stating simply that it was not proceeding "at this time." However, trading platforms such as Kalshi indicate a strong likelihood that Cerebras will revisit its public debut before the following calendar year.
Adding further momentum, recent reporting from Bloomberg suggests that Cerebras Systems is actively pursuing an additional funding round intended to raise around $1 billion. This capital influx would value the enterprise at approximately $22 billion on a pre-investment basis, signaling robust investor confidence in the company’s technological capabilities and future market potential.
Meanwhile, OpenAI continues to strategically enhance its AI portfolio through acquisitions and partnerships that extend beyond computing infrastructure. An acquisition of healthcare technology firm Torch underscores OpenAI’s ambitions to apply artificial intelligence across various industry sectors. Additionally, collaborations such as the recent agreement with SB Energy, affiliated with SoftBank Group, highlight OpenAI’s commitments to fortifying its data center operations through key development and execution partnerships.
Together, these developments reflect a broadening ecosystem in which OpenAI leverages a diverse set of AI hardware providers and development collaborators to scale its capabilities. Within this framework, Cerebras Systems emerges as a pivotal player, poised to challenge established chip manufacturers like Nvidia and AMD by delivering large-scale computing power critical to training and deploying advanced AI models.
The scale of the 750-megawatt computing power agreement represents a significant capital deployment in AI infrastructure, illuminating the substantial hardware demands posed by contemporary AI applications. This partnership may influence competitive dynamics within the semiconductor industry, particularly as Cerebras seeks to capitalize on increased market demand and prepare for future financial events such as its anticipated public offering.
While financial specifics and operational details remain limited in disclosure, this alliance between OpenAI and Cerebras reinforces the critical importance of cutting-edge hardware solutions necessary to maintain leadership in artificial intelligence innovation. Continued developments in this domain will be closely observed by market participants monitoring technology advancements and investment opportunities.