Cerebras Systems' stock has fallen to its lowest level since the company went public, capping a steep slide that began almost immediately after its IPO.

The decline puts fresh pressure on the most prominent challenger to Nvidia's dominance in AI accelerators. Cerebras built its business around wafer-scale processors — chips fabricated as a single, massive die spanning an entire wafer rather than a small section of one. That architecture, manufactured on TSMC process technology, was designed to deliver far more compute per chip than conventional GPUs.

Nvidia's grip on the market is the central problem. The Santa Clara giant controls the overwhelming share of AI training and inference hardware sold today, and its roadmap of successive accelerator generations has made it harder for any rival — startup or incumbent — to win design slots at major AI labs and cloud providers.

Cerebras is not the only casualty. Investors have grown more discriminating across the AI hardware sector, punishing companies whose valuations outran their revenue or whose competitive position against Nvidia looks uncertain. A post-IPO drawdown to record lows signals that the market now discounts the startup's ability to convert its architectural differentiation into sustained commercial traction.

The stock's slide matters beyond one company's market capitalization. Cerebras was among the most closely watched attempts to break Nvidia's hold on AI silicon, and its post-IPO performance will shape how public markets price future AI chip listings. A weak debut cycle could push other accelerator startups toward acquisition or private funding rather than their own offerings.

For now, Cerebras must prove that wafer-scale computing can win paying customers at scale in a market where Nvidia sets the pace on performance, software ecosystem, and supply. Unless the company can point to concrete order momentum, the stock's path back toward its IPO price looks difficult.