Key Points

• Cerebras Stock price rises 17% after stronger earnings and guidance are announced
• Microsoft joins the Cerebras customer list, and AI cloud revenue grows by 272%
• $26.1 billion of backlogged orders highlight growth potential, but also carry execution risks

This round of gains occurred after AI stock went through a difficult period, during which NVIDIA's leadership crisis raised questions about the industry's future from the outside world. The performance of Cerebras shows that despite the uncertainties in the overall market, there is still strong demand for alternative AI computing technologies.

Cerebras Revenue grows by 70%, Microsoft joins the list of AI cloud customers

According to the financial report released by the company on November 12, core revenue in the third quarter reached $230.9 million, an increase of 70% compared to the same period last year.

Core cloud and service revenue soared by 272% to $144.9 million, reflecting the rapidly growing demand for high-speed AI inference.

Revenues calculated using the GAAP methodology were lower, at $210.8 million; the company recorded a net loss of $161.2 million. The difference between the reported figures and the core methodology reflects adjustments that include customer warrants and certain data center costs.

Microsoft has become a new super-large-scale customer, joining the continuously expanding network of corporate and cloud partnerships of Cerebras.

This collaboration adds another potential competitor to NVIDIA's ecosystem based on GPU, although Cerebras has not disclosed the financial value of its contract with Microsoft.

This progress comes at a time when AMD and Broadcom are gaining momentum in the field of replacing AI hardware.

Cerebras achieves differentiation through wafer-level processors, which are specifically designed for high-speed inference and are particularly suitable for scenarios that require rapid response from large AI models.

Cerebras More than $26.1 billion in backlogged orders are mostly related to OpenAI.

As of September 30, the remaining performance obligations disclosed by Cerebras amounted to $26.1 billion, which is higher than the approximately $25.4 billion in the previous quarter.

However, backlogged orders do not equate to cash received, nor do they equate to confirmed revenue. Some of these obligations also include estimated costs transferred from data centers; therefore, this figure is different from future operating profits.

This distinction is very important, as AI infrastructure financing increasingly relies on long-term customer commitments to support large-scale upfront investments.

Cerebras has secured a total data center capacity of 700 megawatts, with delivery planned by the end of 2027. At the same time, it is also expanding its manufacturing capabilities through Flex and Sanmina.

The company's collaboration with Oracle will also bring reasoning driven by Cerebras to Oracle Cloud Infrastructure in 2027, providing enterprise customers with another access path.

This expansion is similar to the aggressive capacity-building strategies adopted by Nebius and CoreWeave. Under these strategies, strong demand must be sufficient to justify large-scale infrastructure investments.

Cerebras Upgrades 2026 Outlook: Intensified Competition in NVIDIA

Cerebras has raised its annual core revenue forecast from the previous $880 million to $890 million to $940 million to $960 million.

For the fourth quarter, management expects core revenue to be between $310 million and $330 million, with the core operating margin ranging between -1% and 1%.

CEO Andrew Feldman also stated that the company expects its revenue to triple by 2027, although this goal depends on whether the computing power capacity can be met.

This uncertainty has already affected competitors. Nebius recently downgraded its outlooks because construction delays have pushed back the expected production capacity to 2027.

The growth of Cerebras also coincides with a time when the stock valuation of AI is under increasing scrutiny due to its profitability and financing costs.

Emir Abyazov

Editor-in-Chief Coinpaper, responsible for driving data-driven editorial operations, implementing a discovery mechanism guided by SEO, and creating narratives that resonate with the audience, covering the fields of cryptocurrencies, AI, and fintech.