The Hardware Bottleneck AI Can’t Fix
Hardware engineering is stuck in a pre-software era, where testing and data management lag decades behind software's rapid iteration cycles. Jason Haack, CEO of Nominal, explains that while software teams enjoy instant feedback loops, cloud elasticity, and real-time observability, hardware teams still rely on physical hard drives, manual data transfers, and weeks-long debugging cycles. The result? A 'hardware bottleneck' that slows innovation despite massive investments in AI and automation. Nominal’s platform tackles this by unifying the entire hardware data supply chain—from high-frequency sensor data and video feeds to simulation results—into a single, real-time system that supports both live control room monitoring and deep post-test analysis. The real breakthrough isn’t just tooling; it’s rethinking how we treat physical assets as 'pets' rather than 'cattle,' with rigorous versioning, event tagging, and lineage tracking. Yet, even as AI agents transform software development, they remain largely absent in hardware—partly because physical tests are slow, expensive, and safety-critical. The future, Haack argues, lies in using AI not to replace engineers, but to empower them: enabling intent-based UIs, automating data tagging, and creating feedback loops that let agents learn from real-world test data. But until we solve the data infrastructure first, AI in hardware will remain a distant dream.
Hardware testing still relies on physical hard drives and manual data transfer, with feedback loops that can take days—orders of magnitude slower than software's instant CI/CD.
Nominal’s platform unifies high-frequency sensor data, video feeds, and simulation results into a single real-time system, enabling both live control room monitoring and deep post-test analysis.
Physical hardware assets are treated as 'pets,' not 'cattle'—each one is unique, expensive, and mission-critical, requiring rigorous versioning, event tagging, and lineage tracking.
AI agents haven’t transformed hardware engineering because physical tests are slow, expensive, and safety-critical; feedback loops are too long to support autonomous iteration.
The future of hardware innovation lies in using AI to empower engineers—not replace them—through intent-based UIs, automated data tagging, and feedback loops that learn from real-world test data.
…and 3 more takeaways available in PodZeus
The Hardware-Software Divide
Software engineering has matured with powerful tooling for observability, CI/CD, and data management. Hardware engineering, however, still operates in a pre-digital era, with slow feedback loops and fragmented data workflows.
Introducing Nominal: The Data Platform for Hardware
Jason Haack, CEO of Nominal, explains how his background in distributed data systems at Palantir and cloud infrastructure at Vercel led to building a platform that helps hardware teams move at software speed.
The Cost of Slow Feedback Loops
Hardware testing is expensive and slow—physical test beds cost millions, and data retrieval can take days. This creates a bottleneck that prevents rapid iteration and innovation.
From Hard Drives to Streaming Data
The shift from batch data collection (hard drive transfer) to real-time streaming (e.g., Starlink) requires new architectural approaches. Nominal helps customers navigate this transition seamlessly.
The Architecture of High-Frequency Data
Hardware generates massive volumes of time-series data—up to a million data points per second. Nominal’s system supports both low-latency control room dashboards and cold storage for deep analysis.
“Not yet, not yet. Let's start with video games and then we can move on to physical systems.”
“Our users, they lose sleep because they are like, what if the data has a story being told in it and I just am not seeing it? And that's the thing that makes or breaks this. asset or system that we're developing.”
“But that requires this concept of like, well, the hardware can be tested in a unit testable manner. And that feedback loop can then introduce the opportunity for something like a design agent to understand what's working and not working in a hardware design.”
Host
Guest
Nominal
organization
Jason Haack
person
SpaceX
organization
Tesla
organization
Palantir
organization
Fidelity
organization
Vercel
organization
Air Force
organization
NASA
organization
Tiger Data
organization
The Hardware Bottleneck AI Can’t Fix
52m • 6/2/2026
Tech Whistleblower: You Only Have 3 Years Left Before This Hits! - Mo Gawdat
2h 1m • 6/1/2026
Is Waymo's Lead Becoming Insurmountable?
40m • 5/30/2026
HYPE Makes History While Bitcoin Keeps Bleeding #CryptoTownHall
52m • 6/1/2026
Market Matters: Shay Boloor on Next AI Boom, Recapping PLTR & DOCN Investments
41m • 6/1/2026
Start discovering podcast insights today
Start with a 7-day trial and explore a growing catalog of popular podcasts. No credit card required.
No credit card required • 7-day trial • Cancel anytime