TechCrunch Disrupt 2026 Roundtables Highlight Physical AI and Data Infrastructure

TechCrunch Disrupt 2026 opens in San Francisco next week, gathering more than ten thousand founders, investors, and operators for three days of deep‑dive conversations. The event’s roundtable program is designed for hands‑on exchange rather than passive listening, giving participants direct access to practitioners tackling the same infrastructure challenges we see in GPU‑accelerated workloads.

Key Themes

• Physical AI data pipelines – edge cases, multimodal ingestion, and reliable real‑world deployment.
• Specialized post‑training datasets – where high‑value data originates and how it fuels the next model generation.
• Enterprise software orchestration – why AI exposes fragmentation and what orchestration layers will replace traditional SaaS stacks.
• Legacy industry modernization – opportunities in construction, energy, and infrastructure where compute meets the physical world.

Why it matters for GPU / AI infrastructure: Each theme hinges on massive, low‑latency compute. Physical AI and robotics demand GPU clusters at the edge; specialized data workloads require high‑throughput training farms; and new orchestration layers will run on accelerated hardware. Understanding these trends helps capacity planners size GPU clouds, choose interconnects, and design scheduling policies that match real‑world demand.

If you are shaping GPU capacity or building AI‑native services, the Disrupt roundtables offer a concise view of where the market is heading and which workloads will drive the next wave of infrastructure investment.