Cerebras Files for Nasdaq IPO Targeting $35B Valuation

Cerebras Files S-1 for Nasdaq Listing

Cerebras Systems, the AI chip startup behind the world's largest computer chip, filed its S-1 registration statement on April 17, 2026, formally kicking off its path to a public listing on the Nasdaq under the ticker symbol CBRS. The company is targeting a valuation of approximately $35 billion and plans to raise around $3 billion in the offering, with Morgan Stanley serving as lead underwriter.

The filing represents a 52% premium over Cerebras' last private valuation of $23.1 billion from its Series H round completed in February 2026. If successful, Cerebras would become one of the largest AI-focused IPOs in history and the first pure-play AI chip company to go public since the current wave of AI infrastructure investment began.

Revenue Growth and Customer Base

Cerebras has built its business around the WSE-3 (Wafer Scale Engine 3), a chip containing 4 trillion transistors and 900,000 cores that occupies an entire silicon wafer. The company claims performance advantages of 21 times over Nvidia's DGX B200 at one-third the cost and power consumption, a proposition that has attracted major enterprise customers.

The IPO filing is anchored by a massive $20 billion-plus multi-year compute deal with OpenAI, reported the same day as the filing. Oracle has also confirmed running Cerebras hardware for customer workloads on Oracle Cloud Infrastructure, while multiple other hyperscalers and AI labs are evaluating the platform. The breadth of the customer pipeline addresses a key investor concern: that Cerebras was overly dependent on a small number of contracts.

Competing in the Shadow of Nvidia

Cerebras enters the public market at a time when competition in AI chips is intensifying. Nvidia dominates the market with its GPU ecosystem, but the industry's shift from training to inference workloads has opened opportunities for alternative architectures. AI chip startups globally raised $8.3 billion in 2026, with companies like MatX, Ayar Labs, and Etched each securing $500 million rounds.

"The fastest AI hardware for training and inference — that's what we're building, and the market is validating our approach." — Andrew Feldman, CEO, Cerebras Systems

Nvidia has responded to the competitive threat by acquiring inference startup Groq's assets in a $20 billion deal in December and investing $4 billion in photonics companies. The GPU giant's upcoming Vera Rubin platform, set for the second half of 2026, promises further performance gains. But Cerebras' wafer-scale approach offers a fundamentally different architecture that some AI engineers believe is better suited to the scaling demands of next-generation models.

AI IPO Window Opens

Cerebras' filing comes amid growing expectations of an AI IPO wave in 2026. OpenAI is reportedly targeting a public listing as early as Q4 2026 at a potential valuation approaching $1 trillion, while Anthropic — valued at approximately $380 billion — is also weighing a listing. SpaceX has already filed confidentially for its own IPO. Prediction markets currently rank Cerebras as the second most likely tech IPO after SpaceX for the year, according to Benzinga.

The IPO market has been relatively quiet since the AI boom began in earnest, with most major AI companies choosing to remain private and raise massive venture rounds instead. Cerebras' decision to go public now signals confidence that public market investors are ready to value AI infrastructure companies at the premiums previously reserved for private markets.

What This Means for Engineers and Job Seekers

Cerebras going public would create a new publicly traded employer in the AI chip space, potentially expanding hiring as the company scales production and customer support. Engineers with expertise in wafer-scale computing, AI compiler optimization, and systems software are particularly relevant. For AI practitioners more broadly, the growing diversity of chip architectures beyond Nvidia means that hardware-aware model optimization skills — including quantization, kernel tuning, and multi-backend deployment — are becoming essential career differentiators in 2026.