Hardware Podcasts

How FPGAs become the gatekeepers of Physical AI Security

The Robot Report Podcast · 4 Sept 2026 · 1h 08m

What this episode covers

Eric Sivertson, VP of Security Business at Lattice Semiconductor, explains how FPGAs provide deterministic guardrails for physical AI and robotics, addressing the shift from traditional IT risk to real-world safety risks. He details how FPGAs can monitor system buses in real-time to enforce cyber resilience, ensuring that probabilistic AI decisions are checked against deterministic hardware constraints. The discussion highlights the vulnerability of main processors like GPUs and MCUs, which lack internal non-volatile memory, and positions Lattice's Root of Trust (ROT) components as essential for secure boot and runtime recovery. Sivertson argues that integrating these security features early in the design cycle is critical for humanoids and autonomous systems to gain market trust and avoid regulatory backlash.

Most relevant to

early-stage companies

Topics

Mentioned

Companies
Lattice Semiconductor, Nvidia, Hugging Face, Skilled AI, Pterodine Robotics, Jaka Robotics, Universal Robots, John Deere, Tesla
Chips
FPGA

Related episodes from other shows

Other shows on the same ground, closest first.

  • The Circuit · 37m

    Episode 68: The Wide World of FPGAs

    Esam Elashmawi, Chief Strategy and Marketing Officer of Lattice, explains the unique value of Field-Programmable Gate Arrays (FPGAs) as reprogrammable semiconductor devices that offer time-to-market advantages and differentiation over fixed-function ASICs. He details Lattice's leadership in the small and mid-range FPGA market, highlighting their Nexus and Avon platforms which prioritize power efficiency, performance, and small form factors for applications in communications, computing, industrial, and automotive sectors. Elashmawi discusses how FPGAs enable 'crypto agility' and rapid algorithm updates in AI and security use cases, allowing companies to protect their 'secret sauce' and adapt to vulnerabilities in the field without hardware changes. The conversation also covers Lattice's high barriers to entry, the growing dollar content per server, and specific AI implementations on client devices like laptops for battery management and user experience enhancement.

    Chips, FPGAs & semiconductorsSecurity & privacy
    8 chapters

  • EE Times Current · 41m

    When AI Gets a Body: Physical AI, Humanoids and the Future of Silicon

    Hezi Saar, Executive Director of Product Marketing at Synopsys, discusses the rise of Physical AI, focusing on the semiconductor demands of humanoid robots and autonomous systems. He explains the shift from cloud-only AI to hybrid inference architectures that balance latency, privacy, and cost, highlighting the 'memory wall' bottleneck in edge devices. Saar also details how open standards like MIPI CSE 2.0 and multi-die packaging are critical for reducing risk and enabling secure, scalable physical AI hardware.

    Chips, FPGAs & semiconductorsEmbedded firmware
    8 chapters

  • Moore's Lobby · 1h 01m

    Ep. 28 | How DARPA Stops IC Hardware Hackers in Their Tracks: Insights from Serge Leef

    Serge Leef, DARPA’s Program Manager for the Microsystems Technology Office, discusses his transition from a 30-year career in design automation at companies like Mentor Graphics and Silicon Graphics to leading hardware security initiatives at DARPA. He outlines DARPA’s four attack surfaces for secure silicon—supply chain, side-channel, reverse engineering, and hardware Trojans—and explains the agency's strategy of embedding security into design tools, likening it to fluoridating water. Leef highlights the economic realities of hardware security, noting that while merchant semiconductor companies invest heavily in bespoke protection, IoT manufacturers often ignore it until forced by crisis, and emphasizes the critical vulnerability of untrusted third-party IP.

    Security & privacyEDA & CAD tools
    8 chapters

  • EE Times Current · 23m

    Security for SoC Interfaces Takes Center Stage in Data Protection

    Dana Neustadter, Senior Product Manager for Security Solutions at Synopsys, explains why securing System-on-Chip (SoC) interfaces is now a mandatory requirement rather than an afterthought. She details how threats vary by application, noting that automotive security is inextricably linked to safety, while IoT devices often lack basic firmware integrity checks. Neustadter highlights Synopsys' Secure Interfaces solution, which integrates hardware root-of-trust and encryption into controllers for standards like PCIe, CXL, and DDR, ensuring data confidentiality and integrity across all links. The discussion concludes with the urgent need to prepare for post-quantum cryptography and the geopolitical fragmentation of security standards.

    Security & privacyPCB design & layout
    7 chapters

  • Hackaday Podcast · 1h 08m

    Ep048: Truly Trustworthy Hardware, Glowing Uranium Marbles, Bitstreaming the USB, Chaos of Congress

    Hackaday editors Mike Szczys and Elliot Williams review the week's most interesting hardware hacks, starting with Elliot's experience at the Chaos Communications Congress (36c3), where he showcased a hoverboard-derived electric cargo carrier and discussed various community projects. The hosts then cover a diverse range of builds, including a verifiable hardware platform called 'Be Trusted' designed to prevent tampering, a uranium marble-based random number generator, and a VGA-to-USB adapter reverse-engineered for bitstreaming. They also highlight creative uses of 3D printing for paper pulp molds, a discrete 1-kilopixel camera, and the democratization of FPGAs and AI on the edge.

    Mechanical engineeringSecurity & privacy
    8 chapters

The summary, chapters and topics are generated from the episode’s audio by this directory, not written by The Robot Report Podcast. The episode itself belongs to its publisher.