Flex and Cerebras to Scale CS-3 AI Supercomputer Production 7x

Flex and Cerebras Systems are significantly expanding their manufacturing partnership to scale production of the Cerebras CS-3 AI accelerator system at facilities in Milpitas, California. The collaboration aims to increase CS-3 production capacity approximately sevenfold through 2026, according to details first reported by the companies.

The expansion includes new production lines, additional floor space, advanced test infrastructure, and increased skilled manufacturing workforce at Flex's Milpitas site. The move addresses accelerating demand for high-performance AI infrastructure while demonstrating that sophisticated AI systems can be manufactured domestically rather than relying on overseas supply chains.

Why it matters

This partnership illustrates a broader shift in AI hardware manufacturing. While electronics production has traditionally migrated overseas, the CS-3's unique technical requirements and strategic importance are driving investment in advanced U.S. manufacturing capacity. The expansion also creates high-skilled manufacturing roles in systems integration, quality control, and testing—countering narratives about domestic manufacturing decline in the technology sector.

Manufacturing wafer-scale computing systems

The CS-3 is built on Cerebras' wafer-scale engine architecture, featuring a processor physically larger than conventional AI chips. Manufacturing these systems presents challenges rarely encountered in traditional server production, requiring specialized handling processes, custom tooling, precision calibration, and extensive system-level validation.

"The CS-3 is unlike any computer system ever built, and scaling its production requires an extraordinary manufacturing partner," said Dhiraj Mallick, COO of Cerebras. "People often think the entire AI manufacturing and packaging supply chain lives overseas, but everyday across the U.S., teams of American engineers and technicians are building state-of-the-art AI systems."

Each CS-3 system integrates advanced liquid cooling, high-density power delivery, precision mechanical assembly, and coordinated networking infrastructure. Flex engineers collaborated with Cerebras to develop dedicated assembly flows, automated test stations, and manufacturing methodologies specifically tailored to wafer-scale computing.

Scaling production infrastructure

The Milpitas facility expansion encompasses precision mechanical assembly, high-power electrical integration, liquid cooling installation, optical networking validation, and full-rack system qualification. The site has evolved into a high-throughput manufacturing environment with parallel integration lines, enhanced burn-in and validation areas, and increased warehouse capacity for critical components and finished systems.

Rob Campbell, President of Communication, Enterprise and Cloud at Flex, noted that "every stage of the manufacturing process—from mechanical integration to thermal validation and final system qualification—required deep collaboration between our engineering teams."

Tooling and fixtures will enable multiple CS-3 systems to move through integration and testing simultaneously, expected to significantly increase throughput while maintaining quality and reliability standards required for large-scale AI deployments. Customer demand is coming from AI model developers, cloud providers, and enterprise customers.

Details of the manufacturing expansion were announced by Flex and Cerebras Systems on July 9, 2026.