Overview 

Stop by Cadence Booth #521 and discover how to accelerate AI infrastructure innovation. As AI factories scale, design teams face growing challenges in integration, optimization, performance, and power efficiency. See how the Cadence Reality Digital Twin Platform, AI-driven design methodologies, AI coprocessor (AICP) technologies,
and a comprehensive IP portfolio help reduce design complexity, speed time to market, and optimize AI workloads across AI accelerators, DSPs, and edge intelligence platforms. Meet our experts, explore real-world solutions, and learn how to build more efficient, scalable AI infrastructure with confidence. 

Cadence Presentations at AI Infra Summit

Session Title: The Next Generation of Tensilica Extensible DSP for AI

Date: Wednesday, September 16 | 12:00pm-12:20pm

Track: Compute

Speaker: James Kim, Chief Architect, Cadence

Summary: As AI continues to move to the edge, system architects face growing demands for higher performance, software portability, and development efficiency. At the same time, increasingly complex workloads are driving the evolution of processor platforms and programmable DSP architectures that support a wide range of signal processing and AI applications. This session explores key trends shaping next-generation edgecomputeplatforms, including architectural flexibility, open ecosystems, and software scalability.Itdiscusses how these innovations are helping designers build more capable and future-ready AI systems.

Session Title: Scaling Physical AI: Five Questions Worth Asking

Date: Wednesday, September 16 | 3:50pm-4:10pm

Track: Physical AI

Speaker: Frank Ham, R&D Fellow, Physical System Design and Analysis

Summary: Physical AI has scaled its training stack remarkably fast. Compute, parallel simulation, policy optimization, and synthetic data have all come a long way in just a few years. What has not kept pace is our ability to say, with rigor, that a policy is ready to operate a real machine, in a real building, for a real duty cycle.

This session offers five open questions rather than a solution. We ask them from the perspective of a company that builds computational software for the design, manufacture, and operation of electronic, mechatronic, and mechanical systems — the tools engineers use to decide whether hardware will actually work before it exists.

  1. What does sign-off look like for Physical AI?
  2. Is sim-to-real real?
  3. What will it take for high-fidelity physics to reach back into RL?
  4. Does validation expire?
  5. Do we have the right design environments?

We don't claim clean answers, but we'll take each in turn, drawing on adjacent experiences — from the evolution of sign-off in EDA, to aerospace's long and uneven history with certification by analysis, and ultimately what it takes to build trust when the consequences are physical.

Booth Demos

Stop by our booth to see the latest technology in action. Learn how you can: 

  • Deploy generative AI workloads with higher performance, lower power consumption, and greater efficiency, and see real-time AI image generation at the edge through Vedya Labs’ Stable Diffusion demonstration with Cadence Tensilica Vision DSPs for Imaging and Vision 
  • Accelerate AI Audio Innovation with Generative Deep Noise Reduction on Tensilica HiFi
  • Power the next wave of AI infrastructure with Cadence HBM4E and UALink solutions for higher bandwidth, lower latency, and greater scale 
  • Build smarter AI factories with physics-based simulation to optimize tokens-per-watt performance with digital twins 
  • Get right-sized AI inference and scalability with chiplets and see the Cadence chiplet framework live on our silicon-proven base system chiplet