Reimagining 3D IC Design with AI

EE Times
EE TimesMay 28, 2026

Why It Matters

AI‑driven 3D IC design shortens development cycles and unlocks the performance needed for next‑generation AI workloads, giving adopters a decisive competitive edge.

Key Takeaways

  • AI accelerates 3D IC design space exploration, reducing years to months
  • Thermal, power, and reliability challenges dominate 3D IC development
  • Reinforcement learning optimizes stacking configurations for performance, power, area
  • Global R&D teams in Egypt, Armenia, US drive AI‑enabled tools
  • Data integration across workflows is critical for AI‑co‑pilot effectiveness

Summary

In this Silicon Grapevine episode, Seaman’s Digital Industries software’s Sudashan Deo discusses how artificial intelligence is reshaping 3D integrated‑circuit (3D IC) design. Deo traces his journey from computer‑science graduate to leading R&D manager across Intel, Synopsys, Cadence and now Siemens, highlighting the global, multi‑site team that fuels the effort.

He explains that 3D ICs stack multiple active dies, shortening interconnects to boost performance, improve power efficiency, and shrink form‑factor. However, the approach introduces severe thermal hotspots, power‑integrity issues, mechanical stress, testability hurdles, and massive data volumes that strain traditional EDA flows.

AI enters as a “co‑pilot,” using reinforcement learning and intelligent optimization to explore the combinatorial design space, identify optimal die‑stack configurations, and reconcile thermal, electrical, and mechanical constraints. Deo likens the architecture to a multi‑storey building and AI to a fast lift that moves data between floors, dramatically cutting design cycles from years to months.

The implications are clear: AI‑augmented 3D IC tools promise faster time‑to‑market, lower development costs, and the ability to meet the compute demands of modern AI workloads, positioning chipmakers that adopt these methods at a strategic advantage.

Original Description

As AI pushes for higher computing performance and bandwidth, 3D ICs are unlocking unprecedented performance, yet with exponentially increasing system-level complexity. From tightly coupled thermal, electrical, and mechanical interactions to massive design space exploration, traditional monolithic 2D SoC design approaches are no longer enough.
In this episode, we explore how AI is fundamentally reshaping both chip architectures and the tools used to design them. Siemens' expert Sudarshan Deo will share the latest use cases where AI-driven EDA transforms 3D IC development cycles, enabling engineers to simulate faster, explore more, and deliver 3D IC systems with greater reliability.
🎙 Listen to EE Times On Air podcasts on power, embedded, AI, and more: https://www.eetimes.com/podcasts/?utm_source=eetimes_youtube&utm_medium=social&utm_campaign=description
🎥 Don't miss an episode and subscribe to our YouTube channel: https://www.youtube.com/@eetimes747
Stay up to date with the latest news in the electronics industry:
🎉 Interested in advertising opportunities? Reach out to sales@aspencore.com

Comments

Want to join the conversation?

Loading comments...