By Anhata Rooprai and Max A. Cherney

Aug 12 (Reuters) – Cerebras Systems raised its annual revenue and gross margin forecasts on Wednesday, buoyed by robust demand for its chips from ‌companies ramping up data-center capacity to power AI services.

Still, its shares were down more ‌than 14% in extended trading after closing up 11.6% in the regular session. The stock has gained 15.5% week-to-date.

The chip designer ​is banking on growing demand for inference, the data crunching that occurs when a user queries a chatbot, as it seeks to challenge Nvidia’s dominance in the AI processor market.

Cerebras’ flagship wafer-scale engine (WSE) is a single chip the size of a dinner plate containing trillions of transistors, a design that it says is ‌more efficient than connecting thousands of ⁠smaller graphics processors together, as Nvidia does.

By placing memory directly on the chip, the WSE is built to accelerate inference and reduce the data-transfer delays associated with ⁠conventional graphics processors that rely on separate high-bandwidth memory.

Placing memory directly on the chip has lessened the impact of surging memory prices and placed it in a better position to compete with Nvidia, Cerebras CEO Andrew ​Feldman ​told Reuters in an interview.

“Nvidia’s prices have gone through ​the roof because of HBM prices,” Feldman ‌said, referring to the high-bandwidth memory included with AI processors. “This is a battleground, and if they can’t deliver or they’re having significant component price increases, of course that helps.”

The Sunnyvale, California-based company expects 2026 adjusted revenue between $880 million and $890 million, higher than its previous forecast of $855 million to $865 million.

“We have made rapid progress in key areas required to deliver exceptional growth against our remaining performance obligations of $25.4 billion (contract ‌revenue expected to be recognized in the future), and plan ​to more than triple revenue in 2027,” finance chief Bob ​Komin said.

Annual adjusted gross margin is forecast ​at 41% to 43%, up from 38% to 41% projected earlier. Analysts, on ‌average, estimate 35.89%, according to data compiled by ​LSEG.

Second-quarter sales rose 74.3% ​to $180.11 million. Adjusted loss was $6.91 million, narrower than the $40.5-million loss a year ago.

Cerebras is racing to expand chip volumes to support a $20 billion multi-year agreement to provide AI compute to OpenAI, ​a deal viewed as key to ‌justifying its valuation.

Its core cloud and services revenue, which reflects the OpenAI ramp, nearly ​quadrupled to $127.73 million in the second quarter.

(Reporting by Anhata Rooprai in Bengaluru and Max ​A. Cherney in San Francisco; Editing by Shilpi Majumdar)