As AI systems become more capable, the infrastructure behind them is becoming just as important as the models themselves. At TechCrunch Disrupt 2026, Cerebras Systems CEO and co-founder Andrew Feldman will explore one of the industry’s defining questions: can AI keep scaling?
The session will focus on the rising demand for compute power, energy, and specialized infrastructure as AI adoption accelerates. These challenges are significant, but they are also driving a wave of innovation in chip design, data center architecture, and efficiency-focused AI systems.
Why this matters
Cerebras has built its reputation around rethinking AI hardware, including large-scale chip designs intended to handle demanding workloads differently from conventional approaches. Feldman’s perspective could offer valuable insight into how the industry may continue expanding AI capabilities without simply relying on ever-larger clusters and higher energy use.
- Scaling challenge: AI growth is pushing current hardware and energy systems to their limits.
- Hardware innovation: New compute architectures could unlock more efficient model training and inference.
- Industry impact: Better infrastructure can make advanced AI more accessible, reliable, and sustainable.
While this is a conference preview rather than a product launch, it reflects a positive and important trend: AI leaders are actively working on the next generation of infrastructure needed to support continued progress.