Cerebras Systems is dramatically expanding its manufacturing capacity for the CS-3 AI supercomputer, with a new partnership with Flex expected to increase production sevenfold through 2026. The expansion comes as demand for high-performance AI infrastructure accelerates globally, and it represents a significant commitment to building advanced computing systems domestically in Silicon Valley rather than relying on overseas supply chains.

What Is the Cerebras CS-3 and Why Does It Matter?

The Cerebras CS-3 is built on the company's Wafer-Scale Engine 3 (WSE-3), a processor that is physically larger than any conventional AI chip. Unlike traditional graphics processing units (GPUs) that handle AI workloads, the WSE-3 is designed specifically for large-scale AI training and inference, delivering inference speeds up to 15 times faster than leading GPU-based solutions on open-source models, according to Cerebras' benchmarks. The system integrates advanced liquid cooling, high-density power delivery, precision mechanical assembly, and tightly coordinated networking infrastructure into a single platform.

Manufacturing the CS-3 presents extraordinary challenges rarely encountered in traditional server production. Each system requires specialized handling processes, custom tooling, precision calibration, and extensive system-level validation that goes far beyond conventional computing hardware assembly.

How Is Cerebras Scaling Production in the United States?

Flex, a global manufacturing partner, is expanding dedicated operations for Cerebras in Milpitas, California, with multiple new assembly and integration lines coming online through 2026. The manufacturing footprint devoted to CS-3 production is expected to grow substantially as production accelerates to meet customer demand from AI model developers, cloud providers, and enterprise customers.

Inside the Milpitas facility, production operations span several specialized areas:

• Precision Mechanical Assembly: Careful construction of the physical components and structural elements of each CS-3 system.
• High-Power Electrical Integration: Installation and testing of power delivery systems that handle the substantial energy requirements of the wafer-scale processor.
• Liquid Cooling Installation: Integration of advanced cooling systems necessary to manage heat from the densely packed processor.
• Optical Networking Validation: Testing of the high-speed networking infrastructure that connects the system to data centers and other equipment.
• Full-Rack System Qualification: Comprehensive testing of the complete system before delivery to customers.

The expansion is also contributing to growth in high-skilled manufacturing roles across the region, including positions in manufacturing, systems integration, quality assurance, supply chain management, and testing.

"The CS-3 is unlike any computer system ever built, and scaling its production requires an extraordinary manufacturing partner. Flex brings the technical depth, operational rigor, and manufacturing expertise needed to support that scale," said Dhiraj Mallick, Chief Operating Officer of Cerebras.

Dhiraj Mallick, Chief Operating Officer at Cerebras Systems

What Does This Manufacturing Deal Mean for Cerebras' Business?

Cerebras reported first-quarter revenue of $193.4 million, more than doubling year-over-year, demonstrating strong demand for its AI hardware. However, the company still posted a net loss of approximately $14 million and a pretax margin near negative 6.5 percent, reflecting that Cerebras remains in a scaling phase rather than operating as a profitable business.

The Flex manufacturing expansion is a critical step toward meeting that demand. By increasing CS-3 production capacity sevenfold, Cerebras aims to serve growing orders from hyperscalers and AI infrastructure providers. The company is also planning to build approximately 200 megawatts of AI compute capacity in Europe by late 2027, with initial data centers in France and the Nordic region supporting OpenAI workloads beginning in late 2026.

Stock market reaction to the manufacturing announcement was swift. Cerebras shares jumped approximately 10 percent on the Flex expansion news, reflecting investor confidence in the company's ability to scale production and capture market share in the rapidly growing AI infrastructure sector.

What Are the Key Challenges Ahead for Cerebras?

Despite the positive momentum, Cerebras faces significant execution risks. The company's free cash flow was negative $120 million in the first quarter, driven by heavy capital expenditures and inventory buildup required to support manufacturing expansion. Additionally, Cerebras depends on wafer supply from Taiwan Semiconductor Manufacturing Company (TSMC), creating potential supply chain vulnerabilities.

The company also missed earnings per share expectations in the first quarter despite beating revenue forecasts, triggering a sharp 16 to 17 percent stock selloff on heavy trading volume. This volatility underscores that investor sentiment around Cerebras can shift rapidly based on execution results and profitability milestones.

"The CS-3 does not resemble a conventional server or rack-scale compute platform. Every stage of the manufacturing process, from mechanical integration to thermal validation and final system qualification, required deep collaboration between our engineering teams," explained Rob Campbell, President of Communication, Enterprise and Cloud at Flex.

Rob Campbell, President of Communication, Enterprise and Cloud at Flex

Cerebras' path forward depends on successfully ramping production while managing cash burn, securing adequate TSMC wafer allocation, and delivering on its roadmap for future products including the WSE-4 processor and expanded AI inference capabilities. The company's ability to scale manufacturing in the United States while maintaining quality standards for these extraordinarily complex systems will be closely watched by investors and industry analysts as a bellwether for American advanced manufacturing in the AI era.