Key Highlights
- Cerebras (CBRS) stock declined approximately 10% during its second trading session after an impressive 68% surge on Thursday’s IPO debut, which priced shares at $185 and concluded at $311.07.
- The AI chipmaker secured $5.55 billion through its initial public offering, selling 30 million shares at $185 apiece before opening at $350.
- Based on projected $510 million in 2025 revenues, CBRS maintains a trailing price-to-sales multiple approaching 200x — more than seven times Nvidia’s current ratio.
- The company’s substantial $24.6 billion order backlog shows extreme concentration risk, with $20 billion linked to one cloud services agreement with OpenAI featuring exclusivity terms and termination provisions for delivery delays.
- D.A. Davidson analyst Gil Luria estimates fair value around $115 per share — approximately $25 billion total — closely aligned with the backlog’s worth.
Cerebras Systems (CBRS) stock experienced a significant pullback of roughly 10% on Friday, just one trading day following its remarkable public market debut, as market participants grappled with astronomical valuation metrics versus concerns regarding technological limitations and customer diversification.
CBRS launched at $185 per share on Thursday, jumped to an opening price of $350, and momentarily touched $385 before regulatory trading halts were implemented. The inaugural session concluded at $311.07 — representing a 68% single-day appreciation. By Friday’s opening bell, shares had retreated to approximately $279.99.
The public offering generated $5.55 billion in proceeds, marking it among the most substantial AI-sector listings in recent history. Calculating on a fully diluted basis, the company’s market capitalization exceeded $100 billion.
That represents an extraordinary valuation for an enterprise projecting $510 million in revenue for 2025.
At prevailing market prices, CBRS commands a trailing price-to-sales valuation approaching 200x. Nvidia, for perspective, currently trades at approximately 27x sales. Even factoring in Cerebras’ anticipated growth rates, the disparity remains substantial.
Evaluating the Technology Proposition
Cerebras manufactures extraordinarily large AI chips. The company’s premier Wafer-Scale Engine 3 (WSE-3) measures, according to company specifications, 58 times the size of competing GPU processors and executes inference operations up to 15 times more rapidly than Nvidia-powered alternatives.
The performance benefits are demonstrable. However, market observers highlight accompanying limitations.
Gil Luria from D.A. Davidson observed that the processor’s unconventional dimensions have thus far restricted its application to smaller, less sophisticated computational models. Manufacturing challenges surrounding yield rates also persist — consistent production of functional chips at commercial volumes remains unverified for extensive implementations.
“The technology platform may provide superior performance for certain use cases, but it lacks the versatility of existing AI computing infrastructures,” Luria stated.
CEO Andrew Feldman countered these assertions, informing Barron’s that Cerebras currently operates larger models in private environments and anticipates public demonstrations within coming weeks.
Examining Backlog Concentration
Cerebras disclosed a $24.6 billion order backlog as of year-end 2024. Management projects converting approximately $3.7 billion of these commitments into recognized revenue spanning 2026 and 2027.
However, significant concerns exist. Roughly $20 billion of the total backlog — exceeding 80% — originates from one cloud infrastructure agreement with OpenAI.
This arrangement incorporates exclusivity provisions potentially restricting Cerebras’ capacity to engage with competing frontier AI research organizations. Additionally, termination clauses permit contract cancellation should delivery schedules slip, potentially eliminating substantial backlog value.
Luria assigned an approximate valuation of $25 billion to the enterprise — translating to roughly $115 per share — derived from the backlog assessment.
Cerebras maintains relationships with prominent AI industry participants, including OpenAI, Amazon Web Services, Meta, and IBM.
Revenue expanded 96% year-over-year to $171 million in the latest reporting period. Meanwhile, Nvidia’s data center division generated $62.13 billion during its most recent quarter, representing 75% growth.
