Cerebras posts 92% revenue growth after IPO

According to CNBC, Cerebras Systems reported $193.4 million in GAAP revenue for the first quarter, a 94% year-over-year increase. Core revenue (which excludes certain items) reached $191.3 million, up 92% year-over-year, per the company's Q1 earnings release. CNBC reports a net loss per share of $0.22 and a net loss that narrowed to $14 million from $23.9 million (or $0.46 per share) a year earlier. CNBC reports the stock has been volatile since the May IPO: the offering priced at $185, opened at $350, closed the first day at $311.07, and was trading down about 28% from the open at roughly $227 on Tuesday.

According to Cerebras' Q1 release, the company provided Q2 core revenue guidance of approximately $194 million, representing roughly 88% year-over-year growth, and full-year 2026 core revenue guidance in the range of $855 million to $865 million, representing approximately 69% growth at the midpoint. CNBC reports that analysts have highlighted Cerebras' architecture advantage; a Mizuho note cited by CNBC says Cerebras packs many times more SRAM on its chip than some recent Nvidia designs.

Alongside earnings, Cerebras disclosed a multi-year agreement with OpenAI covering 750MW of compute capacity valued at more than $20 billion, and a multi-year partnership with Amazon Web Services to bring Cerebras fast inference to AWS, per the company's Q1 release and secondary coverage. These deals materially expand Cerebras' committed revenue base and embed the company in two of the largest AI infrastructure ecosystems.

Companies that supply AI infrastructure can show rapid top-line growth when demand for large-model training expands, but public-market valuations often amplify short-term stock volatility. Investors and practitioners watching hardware vendors typically weigh revenue growth and guidance against adoption, total cost of ownership, and integration with cloud ecosystems. The SRAM-dense Cerebras architecture offers throughput advantages at certain inference workloads, but requires cost and integration benchmarking against GPU-based alternatives before procurement decisions.

For practitioners: Monitor third-party performance comparisons, memory-architecture trade-offs (SRAM density versus other memory hierarchies), and vendor guidance execution as leading indicators for procurement and benchmarking decisions.