Tony Funderburk is Director of Global Electronics & High Tech PLM Sales at PTC, where provides guidance and support for companies looking to accelerate innovation, manage product complexity, and drive digital transformation through Product Lifecycle Management (PLM) solutions. With more than 25 years of experience in enterprise software sales and leadership, Tony brings deep expertise across the electronics and semiconductor industries, working closely with organizations to improve product development, strengthen collaboration, and build more resilient operations.
TL; DR: Semiconductor companies are investing heavily in AI, but many projects stall because product data is fragmented, poorly governed, or hard to access. The starting point is engineering continuity: a digital thread that connects design, manufacturing, quality, and supply chain data in one traceable, governed foundation.
The semiconductor industry is at the forefront of our AI-defined era. Demand for advanced chips keeps climbing. Product architectures are getting more complex. Companies are using AI to shorten design cycles, improve yield, reduce rework, and make better decisions across the product lifecycle.
AI cannot make up for disconnected product data. When requirements, BOMs, engineering changes, test results, quality records, and supplier information sit in separate systems, AI models lack the context they need to produce reliable outputs.
Before organizations can scale AI, they need to address the hidden cost of disconnected engineering across PLM, MES, ERP, quality, and supplier systems.
AI ambition is outpacing data readiness
That readiness gap is already slowing AI programs.
- Accenture found that 61% of business leaders say their data assets are not ready for generative AI.
- Gartner research found 63% of organizations are unsure whether they have the right data management practices for AI.
For semiconductor organizations, the risk is hard to contain and comes with real business costs. IBM reports that more than a quarter of organizations estimate they lose over $5 million each year because of poor data quality.
In semiconductor companies, those losses show up as rework, delayed decisions, duplicated engineering effort, and slower responses to manufacturing or quality issues.
Innovation is hard, but you can’t afford to wait
Explore our infographic to understand the high price businesses pay when process bottlenecks slow innovation.
Explore the InfographicAI readiness depends on engineering continuity
For semiconductor companies, AI readiness is an engineering continuity challenge. Product portfolios can include tens of thousands of SKUs, deep variant complexity, and products that stay in production for decades. AI needs large data sets and governed product context.
Why Hybrid IDMs Face Semiconductor's Hardest Challenge
Dealing with both internal and external manufacturing entities creates a unique product data challenge.
Read the BlogRequirements, configurations, engineering changes, manufacturing records, test data, quality events, and supplier inputs need to stay connected in ways that preserve meaning and traceability. When they do not, teams spend more time cleaning, stitching, and debating data than using it.
The digital thread creates that connection. It links product data across the full lifecycle, from requirements and design to manufacturing, test, quality, and supply chain. With that context, AI can see what happened, why it happened, and what the downstream impact may be.
How a digital thread enables semiconductor AI
A strong digital thread gives semiconductor teams three advantages:
1. Governed product data.
AI needs clean, versioned, authoritative inputs. A digital thread keeps product data structured and controlled instead of scattered across spreadsheets, duplicated systems, or tribal knowledge.
2. Cross-domain visibility.
Semiconductor development spans EDA toolchains, PLM, MES, ERP, test, and supplier systems. A digital thread connects those domains so AI can learn from lifecycle context instead of isolated snapshots.
3. Knowledge reuse.
Proven IP blocks, validated designs, and characterized process configurations represent years of engineering work. When that knowledge is cataloged and governed, AI can surface reuse opportunities and help teams avoid starting from scratch.
One practical test of that visibility is whether product context remains intact after tapeout, as data moves into foundry, packaging, test, and quality workflows. Explore the post-tapeout coordination challenge.
What semiconductor leaders should prioritize
AI readiness works best as a deliberate sequence; semiconductor leaders should focus on five priorities:
- Audit the data landscape. Identify where critical product data lives, which systems are disconnected, and where traceability gaps create operational risk.
- Establish a governed product data foundation. Use PLM as the system of record for parts, BOMs, configurations, requirements, and engineering changes so AI can work from authoritative data.
- Extend traceability across the lifecycle. Connect design, manufacturing, test, quality, and supply chain data so teams can see the impact of decisions and AI-generated recommendations.
- Build knowledge reuse into the process. Capture validated IP, proven designs, and historical decisions in a searchable, governed structure so AI can help teams reuse what already works.
- Introduce AI incrementally. Start in high-value, data-rich areas such as yield prediction, change impact analysis, intelligent part reuse, or quality anomaly detection. Prove value where the data foundation is strongest, then expand and scale.
The path to AI starts with trusted data
The semiconductor companies that lead in an AI-driven market may not be the ones with the biggest AI roadmaps. They will be the ones with the cleanest, most connected, most trusted product data. Engineering continuity does not slow AI adoption. It makes AI scalable, reliable, and useful.
Ready to strengthen the foundation? Start by identifying where lifecycle continuity breaks today. The most visible gaps often occur between engineering and manufacturing, during external partner handoffs, or across packaging and qualification workflows.
Up Next