Cerebras Systems (NASDAQ: CBRS) had a quarter that demands attention. Revenue hit $193.4 million, a 94% jump from a year ago and $12.7 million above the Street's $180.7 million estimate. The GAAP net loss narrowed to $14.0 million from $23.9 million, though EPS of -$0.22 missed the -$0.14 estimate by $0.08. Those numbers are just the start. The real story is the $20 billion-plus OpenAI deal and a full-year guidance raise that implies 69% revenue growth at the midpoint.
Growth was broad-based. Hardware revenue rose 59% to $110.6 million. Cloud and other services more than doubled to $82.8 million, up 178% year over year. Services now account for 43% of total revenue, up from 30% a year ago. That mix shift is crucial. Services carry higher core gross margins (53% versus 42% for hardware) and are more recurring. The cloud business is becoming the growth engine, not just a hardware add-on.
Gross margins improved. GAAP gross margin hit 44.6%, up from 41.8% a year ago. Core gross margin, which strips out pass-through items and stock compensation, reached 46.5%. Both segments contributed: hardware core margins rose to 42.0% from 30.6%, while cloud services core margins slipped to 52.9% from 69.0%. The services margin decline is a function of scale. As Cerebras deploys more data center capacity, pass-through costs grow, compressing the margin percentage even as absolute profit dollars expand. The company's core operating loss narrowed sharply to $3.5 million from $19.3 million. Adjusted EBITDA turned positive at $12.7 million versus -$15.4 million a year ago.
The OpenAI deal defines the quarter. Cerebras announced a multi-year agreement valued at more than $20 billion for OpenAI to deploy 750 megawatts of Cerebras' high-speed inference compute. The companies also co-launched Codex-Spark, a coding model that delivers over 1,000 tokens per second. This is not a pilot. It is a commercial-scale commitment from the world's most prominent AI company, validating Cerebras' wafer-scale architecture as a viable alternative to Nvidia's GPU clusters for inference workloads. The deal also came with a $1 billion working capital loan from OpenAI in January, which appears on the balance sheet as a $983 million customer loan.
Cerebras also deepened its relationship with AWS, announcing a multi-year partnership for disaggregated inference. The architecture pairs AWS's Trainium 3 chips for prefill with Cerebras' CS-3 for decode. It is a clever positioning: Cerebras is not trying to replace the entire AI stack but rather owning the part where speed matters most.
The capital structure transformed. Cerebras raised $1 billion in a Series H pre-IPO round in February, $1 billion from the OpenAI loan in January, and then $6.4 billion in gross proceeds from its Q2 IPO, the largest semiconductor IPO in history. The balance sheet now holds $3.3 billion in cash, equivalents, restricted cash, and short-term investments, up from $930 million at year-end. The company also closed an $850 million revolving credit facility in April. The cash hoard is earmarked for data center acquisitions to support the OpenAI and AWS deployments. The pace of capital deployment will be a key metric to watch.
Guidance reinforces the momentum. Q2 core revenue is expected at $194 million, up 88% year over year. Full-year core revenue guidance of $855 million to $865 million implies 69% growth at the midpoint. The tight range suggests management has good visibility into the pipeline. The core gross margin outlook of 38% to 41% for the full year is below Q1's 46.5%, likely reflecting the mix shift toward lower-margin hardware as the OpenAI deal scales. Core operating margins are expected to remain negative at -28% to -32%, indicating that Cerebras is still in heavy investment mode.
The beat is real, but the story is the scale of the opportunity. Cerebras has secured the two most important customers in AI (OpenAI and AWS) and raised enough capital to build the infrastructure to serve them. The risk is execution. Data center buildouts are capital-intensive and subject to delays, and the company is still losing money on a GAAP basis. The trajectory is clear. Revenue is doubling. Losses are shrinking. The customer base is concentrated on the leaders. For a company that went public just weeks ago, this is a strong opening hand.