트레이딩키는 세레브라스 시스템스의 2026년 IPO 전망을 제시하며, 상장 전 가치 평가 지표와 투자 타당성을 분석했다.
Cerebras Systems is expected to begin trading on the Nasdaq under the ticker CBRS on May 13–14, 2026, marking the most significant pure-play AI hardware IPO of the year. The company is offering 28 million Class A shares at an expected price of $125 to $135 per share, raised from an initial range of $115 to $125 due to demand exceeding 20 times oversubscription. At the upper end of the revised pricing band, the offering could generate approximately $3.78 billion in proceeds, targeting a total raise of $3.5 to $3.8 billion and establishing a fully diluted valuation of $26 to $27 billion. Morgan Stanley, Citigroup, Barclays, and UBS are leading the syndicate.
Cerebras builds wafer-scale AI chips that fundamentally differ from conventional semiconductor design. While standard chips utilize small dies cut from silicon wafers, Cerebras’s Wafer-Scale Engine (WSE-3) uses the entire wafer as a single chip, measuring roughly 57 times larger than the largest competing GPU die. This architecture delivers vastly more on-chip memory, processing cores, and interconnect bandwidth without the latency inherent in multi-chip GPU clusters. The flagship CS-3 system is engineered for large-scale AI model training and inference, sold as a turnkey compute environment backed by its own software stack. By emphasizing a lower total cost of ownership—requiring fewer systems, less power, and reduced cooling infrastructure—the company has successfully attracted hyperscalers and enterprises. Its most prominent achievement is a multi-year agreement with OpenAI for 750 megawatts of AI compute capacity, representing one of the largest AI infrastructure contracts ever disclosed.
In 2025, Cerebras generated $510 million in revenue, reflecting a 76% year-over-year increase, and posted $238 million in net income. This yields a 47% net margin, a profitability metric rarely seen at this growth stage. At $135 per share, the stock trades at approximately 51 to 53 times its 2025 trailing revenue. By traditional standards, this premium appears steep. However, institutional investors demonstrated strong conviction through the 20x oversubscription at the original pricing tier, indicating a willingness to pay for proven profitability. For context, CoreWeave—the nearest comparable—went public in March 2026 at a $23 billion valuation on GPU cloud revenue and remains unprofitable at a similar scale. Whether Cerebras justifies its valuation post-IPO hinges on three factors: whether revenue growth accelerates beyond 76%, whether gross margins expand as the software and cloud services mix increases, and whether the company secures a second major hyperscaler customer to mitigate concentration risk.
Nvidia currently controls approximately 70% to 80% of the AI accelerator market, leveraging its CUDA software ecosystem to create switching costs measured in years of engineering investment rather than hardware pricing. Cerebras does not aim to displace Nvidia across the board. Instead, it targets defensible niches where its wafer-scale architecture provides inherent advantages: massive model training workloads, serving models requiring extreme memory capacity, and operations where physical footprint is critical, such as sovereign AI initiatives, military applications, or dedicated data centers. AMD’s Instinct MI400 and Google’s TPU v5 will also compete for these tasks. The primary threat to Cerebras’s thesis is if Nvidia continues to narrow the technical performance gap generation by generation while maintaining its entrenched CUDA moat.
Prospective investors should weigh Cerebras’s distinctive strengths against its structural challenges. On the bullish side, the company combines 76% revenue growth with real profitability, tier-one industry validation via the OpenAI partnership, and predictable multi-year revenue streams. Conversely, risks include a highly concentrated customer base—with OpenAI estimated to drive a substantial portion of revenue—a deeply entrenched competitor ecosystem, and a 51x trailing revenue multiple that demands sustained execution. Additionally, a lock-up expiration is scheduled for mid-2026, which may introduce near-term volatility.
Cerebras Systems (CBRS) will list and trade on the Nasdaq beginning May 13–14, 2026. IPO shares priced within the offering range are allocated to institutional investors and qualified retail participants through the underwriter syndicate. Retail investors may purchase shares after trading commences using any online brokerage with Nasdaq access. Investors should verify real-time listing status at nasdaq.com/market-activity/ipos before executing orders.
From an infrastructure intelligence perspective, Cerebras stands out as a top-tier AI IPO: it is profitable, growing rapidly, and already secured one of the largest AI data center deals in history. The 20x oversubscription and upward price revision underscore strong institutional demand. Nevertheless, a 51x trailing revenue multiple requires continuous proof of progress, particularly in sustaining growth momentum, expanding margins, and diversifying its client portfolio beyond OpenAI. Market participants should monitor the opening price relative to the $125–$135 guidance, review the first quarterly earnings post-listing, and track any new hyperscaler agreements. This represents the defining AI chip IPO of 2026.
Disclaimer: The content of this article solely represents the author's personal opinions and does not reflect the official stance of Tradingkey. It should not be considered as investment advice. The article is intended for reference purposes only, and readers should not base any investment decisions solely on its content. Tradingkey bears no responsibility for any trading outcomes resulting from reliance on this article. Furthermore, Tradingkey cannot guarantee the accuracy of the article's content. Before making any investment decisions, it is advisable to consult an independent financial advisor to fully understand the associated risks.