Welcome to the 126th edition of Deep Tech Catalyst, the educational channel from The Scenarionist where science meets venture!
This week, I sat down with Isaac Sigron, Managing Director at Koch Disruptive Technologies. We explored why advanced packaging has become one of the strategic bottlenecks behind AI infrastructure, how the shift from monolithic chips to chiplets changes both technical architecture and venture design, why component companies in AI can still represent very large markets, and how founders should think about pricing, procurement, and team-problem fit when building in complex semiconductor ecosystems.
Key takeaways from the episode:
🧩 Advanced packaging is becoming a core AI infrastructure bottleneck
As transistor scaling becomes harder, the industry can no longer rely on Moore’s Law alone. Advanced packaging is emerging as a way to bring compute, memory, power, and interconnect closer together while addressing yield, bandwidth, and system-level performance constraints.
⚡ The future of compute depends on memory proximity
AI workloads are increasingly limited by the cost of moving data. Bringing memory closer to compute through high-bandwidth memory, chiplets, and eventually 3D stacking can improve bandwidth, reduce power loss, and reshape how performance is designed at the package level.
💼 In AI infrastructure, component plays can become large markets
In AI infrastructure, a single bottleneck can represent a massive opportunity if it sits in the critical path of compute scaling and the team knows how to integrate into the ecosystem.
🤝 Procurement moves slower than technology
Founders introducing new compute architectures often need to frame their business model in terms customers already understand. The best pricing structures are not about extracting more value from the customer, but proving a clear win-win through NRE, per-chip economics, scale, and shared upside.
🎯 Team-problem fit matters as much as product-market fit
In Deep Tech, many teams often attack the same bottleneck at the same time. What separates the most compelling companies is not just the technology, but whether the founding team is uniquely suited to understand the problem, navigate the supply chain, work with customers, and turn technical insight into an investable company.
30 Scale-Up Lessons Learned
30 Execution Lessons Learned from 100+ Deep Tech Founders and Investors.
