By JBizNews Desk
May 11, 2026
Cerebras Systems is preparing to sharply raise both the price and size of its blockbuster initial public offering after investor demand for the artificial intelligence chipmaker overwhelmed Wall Street expectations, underscoring the extraordinary appetite currently driving the global AI infrastructure boom.
The Sunnyvale, California-based company is now considering increasing its IPO pricing range to between $150 and $160 per share, according to two people familiar with the matter who spoke to Reuters on Sunday.
That would represent another major upward revision from the company’s already elevated prior range of $115 to $125 per share.
Cerebras is also expected to expand the number of shares offered to approximately 30 million shares, up from the 28 million originally planned.
At the top end of the revised range, the company would raise roughly $4.8 billion, compared with approximately $3.5 billion under the original structure, implying a fully diluted valuation approaching $32 billion.
The figures remain subject to final pricing adjustments ahead of the expected offering date.
The scale of investor demand has stunned even veteran bankers involved in the transaction.
Orders for the offering have reportedly exceeded available shares by more than 20 times, according to the Reuters report, forcing underwriters to repeatedly revise pricing higher during the roadshow process.
Just days earlier, Bloomberg had reported that Cerebras was already preparing to increase the range to $125 to $135 per share.
The latest proposed increase to $150 to $160 signals that demand continued accelerating even after that revision.
The company is expected to price the offering on May 13 and begin trading shortly afterward on the Nasdaq Global Select Market under the ticker symbol CBRS.
The IPO is being led by Morgan Stanley, Citigroup, Barclays, and UBS Group.
If completed near the top of the revised range, Cerebras would become the largest IPO globally so far in 2026, according to data compiled by Dealogic.
The offering also marks a remarkable turnaround for the company itself.
Cerebras originally attempted to go public in 2024 but withdrew the offering after U.S. regulators launched a national security review tied to investment involvement from the United Arab Emirates.
That review concluded earlier this year, clearing the company to proceed with the current listing.
Now, less than two years later, the same company that could not complete its IPO is poised to become one of the hottest AI-related public offerings in modern market history.
The enthusiasm surrounding Cerebras reflects both broader investor appetite for artificial intelligence infrastructure and the company’s increasingly unique position inside the AI hardware ecosystem.
Unlike traditional semiconductor firms, Cerebras specializes in so-called wafer-scale chips — processors physically much larger than conventional graphics processing units, or GPUs.
The company’s chips are specifically optimized for running advanced artificial intelligence systems at scale.
While Nvidia continues dominating the AI training market, Cerebras has increasingly focused on another rapidly growing segment of the industry: AI inference.
Inference refers to the computational process allowing deployed AI systems to actually respond to user requests in real time — the operational side of artificial intelligence after models are already trained.
As generative AI applications scale globally, many analysts believe inference demand may eventually rival or surpass the enormous spending currently devoted to training large language models.
That shift has positioned Cerebras favorably.
The company has secured major customers including Amazon and OpenAI since withdrawing its original 2024 IPO filing, developments that substantially strengthened investor confidence heading into the offering.
OpenAI alone continues spending at extraordinary levels to support inference capacity powering ChatGPT and related products used by hundreds of millions of people globally.
Meanwhile, Amazon Web Services has been racing to expand AI infrastructure capacity across its cloud platform as enterprise demand accelerates.
The broader spending environment across the technology industry is also fueling enthusiasm for AI infrastructure companies.
Analysts at Morgan Stanley recently projected that the world’s five largest hyperscalers — Alphabet, Amazon, Microsoft, Meta Platforms, and Oracle — will increase artificial intelligence-related capital expenditures by nearly 80% during 2026 to approximately $805 billion.
The bank forecasts that figure could rise further toward $1.1 trillion by 2027.
That spending directly benefits the semiconductor firms, networking providers, memory suppliers, and infrastructure companies powering the AI ecosystem.
Investors increasingly view those businesses as occupying critical bottlenecks inside the global AI supply chain.
The funding environment for artificial intelligence startups remains equally aggressive.
AI companies attracted roughly $24.2 billion in venture capital funding during February 2026 alone, while semiconductor valuations across both public and private markets have surged as investors continue bidding aggressively for exposure to the AI trade.
For Wall Street, Cerebras’s IPO may ultimately symbolize something larger than a single semiconductor company going public.
It illustrates how completely investor psychology surrounding artificial intelligence has transformed in less than two years.
A company unable to complete its IPO in 2024 is now preparing to enter public markets with one of the most heavily oversubscribed offerings of the year.
And judging by the pace of demand, investors still appear willing to pay almost any price for a stake in the infrastructure powering the AI revolution.
— JBizNews Desk
© JBizNews.com. All rights reserved. This article is original reporting by JBizNews Desk. Unauthorized reproduction or redistribution is strictly prohibited.
