PDF Solutions Launches Exensio Aurora for Semiconductor Data

PDF Solutions (Nasdaq: PDFS) has unveiled Exensio Aurora, a new analytics architecture designed to process semiconductor manufacturing data at petabyte scale. The company identifies a critical inefficiency in the industry, noting that less than 5% of generated chipmaking data is currently analyzed. This new platform aims to close that gap by deploying agentic AI across manufacturing operations and the global supply chain.

The launch responds to structural shifts in the semiconductor sector driven by 3D architecture, advanced packaging, and heterogeneous integration. These technologies are dissolving traditional boundaries between front-end and back-end production, dispersing manufacturing globally and creating an explosion of data complexity. Exensio Aurora is purpose-built to handle this volume, leveraging PDF Solutions’ 35 years of domain experience to align and analyze data from across the ecosystem.

Architecture Targets 25x Performance Gains

The platform utilizes a distributed engine designed to deliver approximately 25 times faster performance at comparable hardware costs. This technical approach emphasizes server-side analytics with a thin client, allowing for workflow-based usage and no-code analytics creation. By integrating a natural-language interface, the system lowers the barrier for engineers to interact with complex datasets without requiring extensive programming expertise.

Key components include an AI-ready Manufacturing Data House and integrated model lifecycle management built on Kubernetes for edge deployment. The architecture incorporates LLM-enabled agents to facilitate automated workflows. To mitigate risks associated with generative AI, the platform implements workflow-centric governance designed to reduce AI hallucinations, ensuring that outputs remain grounded in verified manufacturing data.

Beta Access Begins in September

PDF Solutions plans to provide beta access to a limited number of early adopters starting in September 2026. This phased rollout allows the company to refine the platform’s capabilities before broader market availability. The initial release focuses on validating the scalability and performance metrics claimed in the architectural design.

First public demonstrations of Exensio Aurora are scheduled for the PDF Solutions CONNECT 2026 event, taking place on October 15-16 in San Francisco. The event will showcase the platform’s ability to handle petabyte-scale data loads and deploy agentic AI solutions for operational efficiency. Registration for the conference is currently open via the company’s official event portal.

Strategic Alignment With Industry Growth

CEO John Kibarian emphasizes that the growth of hybrid packaging is creating a larger and more complex global supply chain. He states that delivering operational efficiency in this environment requires AI-driven collaboration applied at scale. The company ties this launch to its reaffirmed 2026 growth targets, citing recent Exensio customer wins and higher quarterly revenue as indicators of market demand for advanced data solutions.

According to reporting from GN stocks/chips, the initiative positions PDF Solutions to capitalize on the industry’s shift toward heterogeneous integration. By addressing the structural data bottleneck, the company aims to become a central node in the semiconductor data ecosystem, facilitating collaboration between diverse manufacturing partners.