ESD Alliance schedules a Gen-AI for chip design and security webinar for 10 September 2026
The ESD Alliance's Savage on Security series returns on 10 September 2026, examining how multi-agent LLM systems will automate SoC design and security verification.

The ESD Alliance has announced the next instalment of its Savage on Security webinar series, titled "Gen-AI for Chip Design & Security," scheduled for Thursday, 10 September 2026, at 9 am PDT.[1] The session will examine how multi-agent intelligent assistant systems will be architected to automate and augment SoC design and security verification.[1]
The speakers and their background
The webinar is hosted by Warren Savage, a researcher at the University of Maryland's Applied Research Laboratory for Intelligence and Security (ARLIS).[1] Savage founded IPextreme in 2004, a semiconductor IP commercialisation company that was sold to Silvaco in 2016. Before that, he held technical and executive roles at Fairchild, Tandem Computers, and Synopsys.
Joining him is Mark M. Tehranipoor, Intel Charles E. Young Endowed Chair Professor and chair of the Department of Electrical and Computer Engineering at the University of Florida, where he also directs the Florida Institute for Cybersecurity (FICS) Research. Tehranipoor's group has been among the most active in applying LLMs to hardware security, with recent work spanning emulation-based SoC fuzzing, LLM-assisted security asset identification, and automated assertion generation.
What the session will cover
The central argument is that monolithic LLMs are insufficient for the complexity of modern SoC security work.[1] The webinar will argue that multi-agent architectures - where specialised models collaborate across distinct sub-tasks - are the right structural answer.[1] The University of Florida group's published SV-LLM framework illustrates the approach in practice: it integrates separate agents for verification question answering, security asset identification, threat modelling, test-plan and property generation, vulnerability detection, and simulation-based bug validation, with each agent using a different learning paradigm such as in-context learning, fine-tuning, or retrieval-augmented generation (RAG).
The broader design-automation angle is equally significant. Beyond security, the session will address how LLMs can support architectural reasoning, specification refinement, and design-space exploration - tasks that have historically required senior engineers working across long cycles.
Key themes the webinar is expected to address:
- Multi-agent system architecture for SoC design and verification
- LLM capabilities beyond code generation: architectural reasoning and vulnerability analysis
- Security verification automation from threat modelling through bug validation
- What "trusted silicon" means when AI generates or checks the design
Why this matters for EDA teams
The Savage on Security series sits within the ESD Alliance's broader push to put hardware security on the same footing as functional correctness in the design flow. The June 2026 ESD Alliance Executive Outlook, held at Cadence's San Jose headquarters, had already flagged agentic AI as the dominant near-term inflection point for chip design and verification teams. The September webinar narrows that lens to security specifically - a domain where the cost of a missed vulnerability is measured not in re-spins but in compromised devices in the field.
The session is open to SemiWiki readers. Whether the multi-agent framing translates from research prototypes into production EDA flows - and on what timeline - is the question worth pressing during Q&A.
Written by Electronics Insider's automated desk from the sources above and published automatically. How we work.
Related
Design & EDAETH Zürich and NYU show JEDEC test loops miss worst-case HBM2 power by 23.6%
Researchers at ETH Zürich and NYU tested 36 real HBM2 chips and found JEDEC measurement loops understate peak power by 23.6%, with read ops and data-bus toggling as the main drivers.
15 Oct 2026
Design & EDAFlowAccel's Joel Jorgensen argues chip teams must design their workflows before AI can help them hit tape-out
Joel Jorgensen of FlowAccel argues that AI tools alone won't save a tape-out schedule - teams must first design the cross-team workflows that let those tools compound into real gains.
15 Oct 2026
SemiconductorsETH Zürich and NYU find JEDEC test loops underestimate HBM2 power by up to 35.4%
Researchers at ETH Zürich and NYU tested 36 real HBM2 chips and found JEDEC-specified current loops can underestimate peak power by up to 35.4%, with implications for AI accelerator thermal design.
15 Oct 2026