PDF Solutions, Inc., which is a leading provider of complete data solutions when it comes to semiconductor and electronics ecosystems, on September 9, 2026, announced Exensio Aurora, the new, highly scalable Exensio analytics architecture.
Exensio Aurora is designed to be a petabyte-scale manufacturing data platform for semiconductors and secure deployment of agentic AI in manufacturing and supply chain operations.
The first public demonstrations of the Exensio Aurora technology are scheduled to be at the PDF Solutions CONNECT 2026 event, on October 15-16 in San Francisco, CA.
Scalable Architectures When it Comes to Semiconductor Data Analytics
The Exensio Aurora from PDF Solutions responds to a structural shift in the semiconductor industry. 3D architecture and advanced packaging, as well as heterogeneous integration, are blurring the line between front-end and back-end and expanding production around the world, fuelling innovation and growth in the industry. This complexity leads to an enormous amount of data, but in most cases the experience of PDF is that less than 5% of semiconductor manufacturing data is analyzed via conventional tools.
According to PDF Solutions’ CEO, President, and co-founder, John Kibarian, “The semiconductor industry thrives on innovation, which is fueled by collaboration. The growth of 3D architecture and hybrid packaging is creating a larger and more complex global supply chain. Against this backdrop, delivering the operational efficiency that the industry needs requires a new kind of AI-driven collaboration, which can only be achieved when AI is applied at scale to align and analyze data from across the ecosystem, built for semiconductors.”
The data within the semiconductor manufacturing industry is pretty long and wide, and business intelligence – BI systems are often limited in their capacity to handle it. Many general-purpose cloud data platforms do not have the ability to mix with manufacturing equipment and processes, and they have no semantic model based on the industry, nor the ability to orchestrate actions across business applications and the supply chain.
This kind of combination of factors, as well as the pressure from several of its customers, has driven PDF Solutions to invest in this highly scalable Exensio analytics architecture to power the PDF Solutions platform, starting more than a year ago. It uses PDF’s deep 35-year experience to address the industry’s structural shift.
According to Said Akar, VP, Enterprise Development, PDF Solutions, “The challenge today is not just the volume of data that needs to be analyzed, but also the need to incorporate more automation, more machine learning, and more AI in that analysis. PDF Solutions’ strategy has always been to anticipate the needs of the semiconductor industry by incorporating in its solutions the latest technology advancement available. AI is the foundational guiding principle for the new Exensio Aurora architecture, designed to deliver unprecedented levels of scalability and new modes of user interaction with the application.”
Meet Exensio Aurora
PDF Solutions is the industry’s leading operations platform when it comes to semiconductor manufacturing. The PDF Solutions Platform is differentiated from other offerings in that it has been engineered to incorporate four capabilities that the company believes are critical to effective management of semiconductor operations:
- PDF proprietary equipment that provides differentiated data to drive yield enhancements;
- Connectivity and control of manufacturing equipment throughout fabs and remote sites;
- Business process coordination between enterprise applications and the supply chain;
- Exensio is at the core of the platform so as to deliver advanced analytics and AI to semiconductor manufacturers, test floors, and system houses, powered by a semiconductor-specific ontology, a manufacturing data house, and an AI model development and deployment capacity.
The new architecture of Exensio Aurora is designed to deliver the scale and speed that is needed for online and offline analytics and to address the data explosion from the proliferation of hybrid packaging and globally distributed manufacturing.
But Exensio Aurora is not just the old model with some speed tweaks. It represents a different design philosophy: server-side analytics with a light-weight client, a workflow-based platform, no-code creation of new types of analysis, and a natural-language interface.
Exensio Aurora is based on the innovation of scalability, AI-readiness, and workflows.
- Analytics at Scale The new Exensio Aurora distributed engine is designed to provide 25X faster performance at no additional hardware cost. With the PDF approach, only the raw data required for a particular visualization is brought to the client, while pre-computed analytics are architected to keep pace with petabyte-scale systems. Data size is designed to be limited only by the analytics cluster, and performance is designed to scale linearly with volume growth, whether in width or length.
- Manufacturing Data House. This evolution of the Exensio data and semantic model is designed to be AI-ready for data that is increasingly produced through logical operations rather than directly measured. It is designed to provide enhanced support for 3D and higher-dimensional data, metadata for custom data models, and powerful search – all directly addressing data that is becoming more dimensional, more hierarchical, more geographically distributed, and increasingly virtual.
- Model Lifecycle. Exensio Aurora is a unified platform that helps users build, manage, and govern models throughout their entire lifecycle. The capability is tightly coupled with the Exensio database for training, simulation, and feed-forward deployment and is architected to enable bring-your-own-model workflows. It is built on Kubernetes and is designed for central training with deployment to multiple edge locations such as outsourced assembly and test houses – OSATs, test floors, and edge boxes.
- Agent AI – Exensio Aurora leverages three decades of semiconductor domain expertise and is built with customized LLM-enabled agents to assemble full workflows, with on-premise options where cloud-hosted solutions are not preferred.
- Interactivity by AI first – Exensio Aurora is a separate thin-client UI with an embedded AI assistant that brings analytics functionality to a broader set of users without the weight of legacy tooling.
- Workflows – Workflows are the linguistic and long-term storage of Exensio Aurora – every analytic is a workflow regardless of how it was created, rules are workflows, ML pipelines are workflows, and even batch analytics are workflows. Block-level workflows are meant to house semiconductor-specific content and context, generate best practice playbooks as flows, offer visibility into how an outcome was achieved, and act as safeguards to ensure consistency and prevent AI hallucination.
Availability
A beta release program will allow a small group of early adopters to access Exensio Aurora in September 2026. The September 2026 release of Exensio Aurora technology will incorporate technology from the aforementioned new offerings: Exensio Aurora Scalable Analytics, Exensio Aurora Manufacturing Data House, Exensio Aurora Scalable Data APIs, Exensio Aurora Automated Analytics, Exensio Aurora StudioAI, as well as Exensio Aurora Advanced Feed Forward.