Cerebras Systems (CBRS), the California-based AI chip designer whose wafer-scale processors compete in a market dominated by Nvidia and AMD, is at the center of a sharp selloff after its debut post-IPO earnings report flagged full-year gross margin guidance that fell well short of its first-quarter performance. The stock dropped 17.81% on June 21, 2026, to $186.01, a price that sits just above the 52-week floor of $185.22 and roughly 52% below the 52-week peak of $386.34.
At a Glance
- Price: $186.01, down 17.81% on the session
- 52-week range: $185.22 to $386.34
- Market cap: $49.79 billion
- RSI: 33.71, approaching oversold territory
- 2026 adjusted gross margin guidance: 38% to 41%, versus Q1 actual of 47%
| Price | 186.01 USD |
|---|---|
| Day change | -40.39 (-17.81%) |
| 52-week range | 185.22 – 386.34 |
| Market cap | $49.79B |
| RSI (14) | 33.71 |
| Volume | 15,477,391 |
A Debut Earnings Report That Rattled the Market
Cerebras went public in a closely watched IPO, and its first earnings release as a listed company immediately put pressure on the stock. The central issue is margin compression. Management guided for full-year 2026 adjusted gross margins in a range of 38% to 41%, a meaningful step down from the 47% the company delivered in Q1. Even so, the guidance topped the analyst consensus of 29.58%, which points to how deeply skeptical the Street had already become before the print.
The comparison with peers is unflattering. Nvidia operates in the mid-70% gross margin range, and AMD sits in the mid-50s. Cerebras is structurally disadvantaged on this metric partly because of the physical scale of its chips. Wafer-scale processors are significantly larger than conventional GPU dies, which raises production costs per unit and compresses margins relative to competitors who can spread wafer yield losses across smaller die sizes. Analysts have also flagged a second pressure point: Cerebras is currently renting back its own systems from an existing customer to meet near-term demand while it builds out additional data center capacity. That arrangement adds cost and will weigh on margins until the company brings more of its own infrastructure online.
What the Numbers Say
Valuation
With a market cap of $49.79 billion and no P/E ratio or EPS figures currently available given the company's growth-stage profile, valuation rests almost entirely on forward revenue expectations and deal flow rather than trailing earnings. The stock is now down more than 27% from its IPO debut price, a decline that reflects both the margin guidance disappointment and a broader cooling of enthusiasm for AI-infrastructure names. At $186.01, the stock is trading within cents of its 52-week low, meaning the market has essentially given back the entire post-listing premium.
Momentum and RSI
An RSI reading of 33.71 places Cerebras close to the conventional oversold threshold of 30. That level does not guarantee a reversal, but it does indicate selling pressure that has been extreme and sustained. A stock pressing against its 52-week low with RSI near 30 is a setup that technical traders typically treat as a potential exhaustion point, though the fundamental catalyst here is a specific earnings miss on a closely watched metric, not a general market rotation. Momentum-based signals carry less predictive weight when a structural repricing is underway.
Yield
Cerebras pays no dividend. Capital return is not part of the company's financial profile at this stage, and all resources are directed toward scaling manufacturing capacity and securing long-term customer agreements. Yield-focused investors have no income component to offset the price volatility.
Bull Case
The most compelling argument for the bulls is the quality and scale of the deals Cerebras has announced. The company signed a $20 billion multi-year agreement with OpenAI, and CEO Andrew Feldman confirmed on the post-earnings call that OpenAI's GPT 5.4 model is already running on Cerebras hardware. Under the terms of that deal, OpenAI is set to deploy 750 megawatts of Cerebras semiconductors, which represents a substantial committed revenue runway. On top of that, Amazon Web Services is expected to begin using Cerebras chips in its data centers, with revenue from that relationship anticipated within the next year. Morgan Stanley raised its price target to $273 from $250 following the earnings release, a signal that at least one major institution sees the selloff as an overreaction to near-term margin noise rather than a structural impairment of the growth thesis. TD Cowen pointed to the OpenAI and Amazon deals specifically as the key drivers of long-term value creation.
Bear Case
The bear case is straightforward. Gross margins in the 38% to 41% range, even if they beat the lowball consensus, are roughly half of what Nvidia earns on its products. In a capital-intensive industry where pricing power and scale matter enormously, a persistent margin gap of that magnitude limits earnings potential and raises questions about competitive positioning as Nvidia and AMD continue to iterate on their architectures. The sale-leaseback arrangement on the company's own systems is a short-term fix that adds cost complexity. Meanwhile, the $49.79 billion market cap implies a valuation that still prices in significant future execution on deals that have not yet produced material revenue at scale. Any slip in the OpenAI or AWS rollout timelines would likely extend the selloff.
Competitive Context and the Margin Gap
The margin comparison table below captures where Cerebras stands relative to the two most direct chip-industry reference points.
| Company | 2026 Gross Margin Guidance (Cerebras) or Approximate Range |
|---|---|
| Cerebras Systems (CBRS) | 38% to 41% (2026 guided) |
| Nvidia | Mid-70% range |
| Advanced Micro Devices (AMD) | Mid-50% range |
The gap is not trivial. Cerebras guided above what analysts expected, which suggests the market had already discounted severe margin pressure. The question is whether the company can close that gap over time as it scales production and reduces its reliance on third-party capacity arrangements, or whether wafer-scale chip economics create a structural ceiling that competitors do not face.
CEO Andrew Feldman has been explicit that the AWS relationship represents a new and significant growth vector. Revenue from that partnership is expected to begin flowing within the year, which means the next two to three quarterly reports will be critical checkpoints for validating whether the company's cost structure improves in line with management's implied trajectory.
Frequently Asked Questions
Why did Cerebras stock fall so sharply after earnings?
The company issued 2026 adjusted gross margin guidance of 38% to 41%, well below the 47% it reported in Q1, raising concerns about profitability as it scales. The guidance also highlighted a cost-adding arrangement where Cerebras rents back its own systems from a client to meet demand, which contributed to investor unease.
What is the OpenAI deal worth to Cerebras?
Cerebras announced a $20 billion multi-year agreement with OpenAI. Under the deal, OpenAI is committed to deploying 750 megawatts of Cerebras semiconductors, and the company's CEO confirmed that GPT 5.4 is already running on Cerebras hardware.
How does Cerebras compare to Nvidia on gross margins?
Nvidia operates in the mid-70% gross margin range while Cerebras guided for 38% to 41% for full year 2026. AMD sits in the mid-50% range. The gap reflects differences in chip architecture, production scale and the current stage of Cerebras' manufacturing buildout.
Is Cerebras near its 52-week low?
As of June 21, 2026, Cerebras traded at $186.01, just above its 52-week low of $185.22. The stock has lost more than 27% since its IPO debut and is down sharply from its 52-week high of $386.34.
Where Cerebras Goes From Here
The stock is pressing against its 52-week floor with an RSI near oversold levels, a price target raise from Morgan Stanley, and two of the most significant AI infrastructure contracts in the industry sitting on its books. The margin story is genuinely weak relative to peers, but the revenue pipeline from OpenAI and the pending AWS ramp gives the company concrete catalysts that will resolve within the next few reporting periods. The next quarter will show whether the cost structure begins to recover as the company builds out its own data center capacity, or whether the margin gap widens further. That data, more than any price-target revision, will determine whether $186 proves to be a floor or a waypoint on the way lower.
