Hardware Podcasts

Inside Simberian's AI-Driven Design Automation

OnTrack · 9 Dec 2025 · 58m · Episode 325

AI in practice: 3 real uses of AI described

What this episode covers

Yuriy Shlepnev, founder of Simberian, discusses the integration of Large Language Models (LLMs) into signal integrity (SI) workflows via Simbeor Commander. He contrasts this agentic approach with previous machine learning attempts that relied on training data, arguing that LLMs are better suited for orchestrating deterministic solvers than for optimization or prediction. Shlepnev demonstrates how the AI agent can autonomously create stackups, import complex data from PDFs, and compare designs, effectively acting as an engineer that calls the underlying simulation tools.

AI in practice

Where this conversation describes AI being used to build or deliver a product. Part of the AI in practice tracker.

DesignSimberian Simbeor 2025.01 (Simberian Commander)

An AI agent system autonomously generates PCB stackups and transmission line models from natural language prompts, calculating physical parameters like impedance and attenuation using deterministic solvers rather than just predicting values.

“So generalization of this is probably impossible. It's just kind of right now what happened. It's difficult to follow in different area. But I'm not aware that of EDA software, LLM is actually walking inside, not as a help system, not as a kind of assistant type, but actually running the software.”

Yuriy Shlepnev · Hear it at 1:36

DesignSimberian Simbeor 2025.01 (Simberian Commander)

The AI agent imports stackup data from disparate formats (PDF, Excel) and automatically compares them to identify discrepancies in material models or roughness specifications that cause impedance mismatches.

“So I open a couple of files and impart. To stack up. Uh, and compare. I should start with for 20 gigabit per 2nd. This is CP28 from Wild River technology … Import to stick up and compare stuck up. Structure. And material models.”

Yuriy Shlepnev · Hear it at 45:24

Engineering & simulationSimberian Simbeor 2025.01 (Simberian Commander)

The system uses LLMs to orchestrate deterministic solvers for pre-layout analysis, allowing the AI to plan and execute complex signal integrity tasks like stackup planning and link optimization without manual button-pushing.

“So my opinion is very simple answers. You have to create tools, deterministic tools. Like search engine, like optimization engine, like analysis engine, signal integrity analysis. So that's what we are building. Our AI 4SI project … LLM just orchestrate things, create plan, if you give it a prompt and I'll show you today.”

Yuriy Shlepnev · Hear it at 14:06

Most relevant to

companies scaling up

Topics

Mentioned

Companies
Simberian, Intel, Wild River Technology
Products
Simbeor, Simbeor Commander
Materials
Rogers 303, Megtron 7

Related episodes from other shows

Other shows on the same ground, closest first.

  • EE Times Current · 31m

    How AI is Changing Every Aspect of EDA

    Satish Kumar Balasubrahmanian, who leads product management for custom IC verification at Siemens EDA, explains how the company applies 'Verifiable AI' to accelerate semiconductor design and verification. He details the evolution from brute-force Monte Carlo simulations to AI-driven tools like Solido, which reduce simulation counts by 100x while maintaining high-sigma accuracy for mission-critical applications like automotive. Balasubrahmanian also discusses the integration of generative AI and LLMs to leverage historical design data, improve workforce training, and automate knowledge retrieval, arguing that these tools must be verifiable and robust to be production-ready.

    EDA & CAD toolsChips, FPGAs & semiconductors
    6 chapters

  • The Amp Hour · 1h 07m

    Intelligent Routing with Sergiy Nesterenko

    Sergiy Nesterenko, CEO of Quilter.ai, explains how his company uses generative AI and parallel cloud compute to automate PCB layout from schematics. He discusses the company's pivot from hardware sensors to software, drawing on his experience at SpaceX where manual routing became a critical path bottleneck. Nesterenko details how Quilter handles constraints, validates designs using built-in Maxwell equation solvers, and targets specific fabs like JLCPCB and Oshpark to ensure manufacturability.

    EDA & CAD toolsAI & machine learning
    7 chapters

  • Moore's Lobby · 56m

    Solving the Hard Problems: Combining AI and Physics for PCB Design

    Sergiy Nesterenko, founder and CEO of Quilter, explains how his company uses reinforcement learning and computational physics to automate PCB layout, aiming to create a 'compiler' for electronics that handles the entire design scope. He recounts his early career at SpaceX, including an internship watching paint dry and a subsequent board fire that highlighted the reliance on tribal knowledge and the failure of traditional auto-routers. Nesterenko details Quilter's workflow, which integrates with existing CAD tools like Altium to generate layouts for test boards and complex single-board computers, while acknowledging current limitations with high-voltage or ultra-dense server designs. The discussion covers the technical strategy of using neural networks to interpolate simulation results, the company's focus on iterative complexity scaling, and a 10-year vision for enabling continuous integration in hardware design.

    PCB design & layoutAI & machine learning
    9 chapters

  • EE Times Current · 23m

    Beyond Automation: Self-Verifying AI Agents for Semiconductor and PCB Design

    Amit Gupta, Chief AI Strategy Officer at Siemens EDA, and Tim Costa, VP and GM of Industrial and Computational Engineering at NVIDIA, discuss their collaboration to build self-verifying AI agents for semiconductor and PCB design. They explain how the Fuse EDA AI agent uses NVIDIA's GPU-accelerated libraries and Nemotron models to execute long-running engineering workflows while verifying outputs against physics-based EDA engines. The speakers highlight a specific case study with NVIDIA's silicon team, where the Salido characterization suite achieved a 10x reduction in turnaround time and token costs by automating manual setup and error handling. They outline three milestones for the future: workflow coverage, verified repeatability, and integration into core engineering infrastructure, aiming to cut tape-out schedules in half.

    AI & machine learningChips, FPGAs & semiconductors
    9 chapters

  • EMS@C-LEVEL · 17m

    Context, Tools, and Trust:Missing Ingredients in Many AI Strategies, with Luminovo's Timon Ruban

    Timon Ruban, Co-Founder of Luminovo, discusses the evolution of AI in electronics manufacturing, contrasting early hype with current practical applications enabled by large language models. He explains that successful implementation requires balancing comprehensive data context with specialized tools, such as part search APIs, to verify AI proposals. Ruban highlights Luminovo's focus on extracting structured data from unstructured sources like PDFs and emails, and developing 'agentic workflows' where AI assists human decision-making in supply chain management rather than replacing it.

    AI & machine learningSourcing & supply chain
    6 chapters

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