Cerebras Systems, the innovative AI chip manufacturer known for its impressively large processors, is currently increasing its initial public offering price range to a range of $150-$160 per share. This substantial rise values the company at approximately $32 billion, reflecting a remarkable leap from its previous private valuation of $4 billion in 2024.
What does an oversubscribed IPO indicate? Cerebras plans to officially launch its shares on May 14, and the offering has garnered considerable interest, being oversubscribed by 20 times. This means that investors are eager to purchase twenty times more shares than the company is offering, a clear signal of demand.
Analyzing growth from $4 billion to $32 billion in less than two years, Cerebras previously secured $720 million in funding from parties such as Alpha Wave Global and G42. The company’s valuation has skyrocketed eightfold before it even makes its public debut.
In the AI chip sector, valuations have escalated by 150% year-to-date in 2026, and it is projected that AI capital expenditures will exceed $500 billion by 2028. Cerebras's main product, the CS-3 wafer-scale chip, boasts an extraordinary capability, delivering 125 petaflops of AI compute, equating to 125 quadrillion calculations per second. Such processing power is essential for large language model operations and complex scientific simulations at scale.
What clients are benefiting from Cerebras technology? The company has deployed its cutting-edge technology in over 100 clusters worldwide, counting notable clients like GlaxoSmithKline and the UAE government among its partners.
How does Cerebras compare to Nvidia? In April 2026, Cerebras launched the CS-3 AI100 Inferencing System, claiming that it achieves inference speeds twice as rapid as the Nvidia H100 clusters, while operating at 40% lower costs. Unlike Nvidia's model of networking thousands of smaller GPUs, Cerebras adopts a unique approach by utilizing an entire silicon wafer as a single massive processor. This strategy reduces communication bottlenecks between chips, an essential factor for efficiently training and running large AI models.
What implications does this have for investors? The raising of the price target to $150-$160 puts Cerebras in a rarefied space for tech IPOs. If the company indeed debuts at a $32 billion valuation, it will rank among the largest technology IPOs in recent US history. Investors should closely monitor Cerebras’s revenue trajectory and gross margins once it transitions to a publicly traded company. Wafer-scale chip production is highly expensive, and optimizing yields on such vast silicon is notoriously challenging.
Given the projection of over $500 billion in AI capital expenditures through 2028, significant players like Microsoft, Google, and Amazon have strong motives to diversify their chip supply chains beyond a singular vendor. While Cerebras has garnered investor confidence through established products and customer relations, the extraordinary 8x valuation surge in under two years creates minimal room for error.