Key Takeaways
- Cerebras (CBRS) shares rose 6% Monday following news of potential expedited inclusion in S&P Dow Jones Indices
- The AI chipmaker debuted with a spectacular 68% first-day gain on May 14, launching at $350 versus its $185 IPO price and securing $5.5 billion
- The company’s Wafer-Scale Engine technology purportedly delivers performance up to 15x faster than GPU systems, with certain applications reaching 1,000x speed improvements
- Between 2022 and 2025, revenue skyrocketed approximately 2,000% to $510 million, though 62% originated from one UAE-based university
- Heavy reliance on a limited client base — including OpenAI and Group 42 — represents a significant business vulnerability
Shares of Cerebras Systems (CBRS) advanced 6% during Monday’s session — just its second complete trading day — following news that S&P Dow Jones Indices planned to expedite the AI semiconductor company’s addition to its benchmark index. Extended trading saw the stock gain nearly an additional 2%.
CBRS finished Monday’s regular session at $296.65, trading within a $272.24 to $303.66 intraday band.
The momentum for Cerebras has been immediate. The firm debuted on the Nasdaq Global Select Market May 14, establishing its initial public offering price at $185 per share. Trading commenced at $350 and concluded its inaugural day at $311 — representing a 68% jump — while securing $5.5 billion in capital. This positioned it as 2026’s largest IPO.
Demand for the offering significantly exceeded supply, with orders reaching more than 20 times the available allocation. Following a retreat to $293 Friday, shares rebounded Monday.
What’s Driving Investor Interest
The enthusiasm centers on Cerebras’ core technology. The company’s Wafer-Scale Engine (WSE) measures 58 times the physical size of Nvidia’s B200 processor. The third-generation WSE contains 4 trillion transistors — dwarfing the 208 billion found in Nvidia’s dual-GPU configuration.
According to Cerebras, this architecture enables inference speeds up to 15 times faster than GPU-powered alternatives, with specific use cases achieving 1,000-fold performance gains. Inference represents the stage where AI models produce outputs.
Organizations can leverage the WSE by acquiring platforms for proprietary data centers or accessing capacity via Cerebras Cloud and select third-party cloud infrastructure.
Commercial momentum has been substantial. Revenue expanded approximately 2,000% from 2022 through 2025, concluding at $510 million last year.
The Customer Concentration Challenge
While the financial metrics appear impressive, there’s a significant caveat. One client — Mohamed bin Zayed University of Artificial Intelligence in the UAE — generated 62% of Cerebras’ total revenue last year.
The firm disclosed this concentration risk in its IPO filing, acknowledging that its reliance on this institution, Group 42 Holding Ltd, and OpenAI “subjects us to a number of risks.” Should any major client reduce spending, the financial consequences could be severe.
By comparison, Nvidia serves a broadly diversified customer portfolio including Microsoft and Amazon — built across three decades of operation.
Cerebras launched in 2015. Nvidia’s history extends to 1993. While comparisons between the companies are frequent, substantial differences exist in scale, product portfolio, and customer diversification.
Nvidia reported over $215 billion in revenue for its most recent fiscal year — a 65% year-over-year increase — and its shares have appreciated approximately 1,400% during the past five years.
Academic research also suggests caution with newly public companies. University of Florida finance professor Jay Ritter’s data indicates IPO stocks have underperformed comparable firms by 3.6% annually in the five years following their listing, when excluding first-day returns.
Cerebras currently carries a $64 billion market capitalization. Confirmation of the S&P Dow Jones expedited inclusion would trigger substantial index fund purchases.


