The semiconductor industry is at an inflection point. As AI has the potential to transform semiconductor design and manufacturing, and supply chains grow increasingly complex, the need for robust data infrastructure has never been more critical. At the recent PDF Solutions Users Conference, industry leaders gathered to share insights on navigating this transformation; and one message rang clear: success in the AI era starts with data management, enabled by a platform built specifically for the semiconductor industry.
The Platform Powering the Industry
PDF Solutions has positioned itself as the operations operating system for the semiconductor ecosystem, offering a comprehensive platform spanning three critical areas: characterization and technology development, smart manufacturing and analytics, and supply chain orchestration. The numbers speak to their reach: over 300 manufacturing sites connected, partnerships with the top 6 foundries, and 99.9% uptime across 8 global regions exchanging exabytes of data.
As CEO John Kibarian noted, the company’s longevity comes from consistently “looking at what’s needed for the market as it evolves.” That evolution today centers squarely on data integration, cross-enterprise collaboration, and AI.
AI is a Business Enabler, But Data Comes First
Perhaps the most resonant theme throughout the conference was the primacy of data infrastructure in any AI strategy. Tom Caulfield, Chairman of GlobalFoundries, delivered a reality check that many in the industry are experiencing: “I don’t know a single company that’s trying to implement AI, and they say, ‘oh, my problem is I can’t make the algorithms work.’ All their problems is “I can’t get the data managed the right way so I can use the algorithm.”
This insight underscores a critical shift in thinking. AI isn’t just a tool; it’s a business and profit enabler, but only when built on a foundation of clean, integrated, accessible data.
The Key Commandments of AI Adoption
Michael Campbell, Qualcomm’s Chief Supply Chain Officer, emphasized the importance of PDF Solutions’ ability to “handle all the variations in our test programs” and create “commonality and drive a baseline dataset.” This speaks to three fundamental requirements for AI success:
Data Integration: Connect your teams and data sources with purpose. Bridge MES, EDA, supply chain, and quality systems through API-based connectivity and industry semantic-based alignment.
Data Quality: Ensure your data is clean and standardized. As Intel Foundry’s Aziz Safa outlined, success requires a unified data platform capable of handling vast amounts of high-quality data, supported by immense computational resources and highly skilled teams of data scientists and ML engineers.
Technical Infrastructure: PDF Solutions’ Studio AI exemplifies this approach; built specifically for semiconductor data, it provides a single platform to train, deploy, infer, and monitor AI/ML models while helping data scientists scale using open-source or custom algorithms.
Breaking Down Supply Chain Opacity
One of the conference’s most compelling discussions centered on supply chain visibility. The fabless-to-OSAT handshake has traditionally been one of the most opaque parts of the semiconductor supply chain. PDF Solutions showed how its platform, now enhanced with the recent acquisition of secureWISE, enables secure data exchange and communication between fabless companies, IDMs, and their OSATs, providing visibility into process signatures and quality issues.
ASML’s Eduard Gerhardt shared that they’ve been using SecureWISE for 20 years, with all ASML equipment tools connected to the platform. The key to long term use of secureWISE? Data confidentiality and encryption that protects proprietary customer information while enabling the remote access necessary for modern operations.
From Systems of Record to Systems of Action
The industry is moving beyond simply recording data to acting on it automatically. Ranjan Chatterjee, VP of Smart Factory Solutions, described how automated orchestrations transform systems of record; both within companies and with external partners; into systems of action. This lays the groundwork for agentic AI that automatically responds to data within complex, multi-system processes.
Intel Foundry’s Wes Sparks emphasized the security imperative in this transition: “You need a single solution for remote ops and you have to have advanced security built into that model. You can’t have multiple solutions. It becomes unmanageable very quickly and it actually creates a security issue for you.”
Scaling Beyond Pilots
A recurring challenge discussed at the conference was moving from proof-of-concept to production-scale deployment. Viswa Anakkara Vadakkath posed the critical question: “How do you make sure that the first step is in the direction of that end goal that you’ve set for yourself?”
The answer lies in maintaining momentum from pilots while demonstrating immediate results. Intel Foundry’s Syed Baquar emphasized what customers are looking for: a single UI, easy data access, integrated analytics, and modern, flexible, scalable systems.
Universal Solutions for Diverse Manufacturing
PDF Solutions isn’t just focused on leading-edge logic. As Jonathan Holt, Senior Director of Product Management, explained, their solutions span compound semiconductors, memory, and the many varieties of silicon manufacturing emerging today. With equipment software installed on over 8,000 semiconductor systems this year alone, spanning front-end, back-end, metrology, and lithography, the company has established itself as the neutral software provider for the industry.
앞으로 나아갈 길
The messages from the PDF Solutions Users Conference paint a clear picture of where the semiconductor industry is headed. Success requires integrated data infrastructure, robust cybersecurity, scalable AI platforms, and the technical expertise to tie it all together.
But perhaps most importantly, it requires a fundamental mindset shift: treating data management not as an afterthought to AI initiatives, but as the essential foundation upon which all transformation is built.
As companies race to implement AI and digital transformation, those who heed Caulfield’s advice; focusing first on collecting and formatting data correctly, will find themselves best positioned to leverage the algorithms that follow. In the semiconductor industry’s next chapter, the winners won’t necessarily be those with the best AI models, but those with the best data to feed them.
The 2025 PDF Solutions Users Conference brought together leaders from across the semiconductor ecosystem to share insights on data, analytics, and mission-critical platforms spanning characterization, smart manufacturing, and supply chain orchestration. All the presentations and videos are available on our website here.