Nvidia's stock retreated below the $200 mark in early Monday trading, with shares falling 1.5% to $198.51, underperforming major market indices. This decline reversed part of a 15% advance over the past month, highlighting renewed competitive pressures in the artificial intelligence chip sector.
The immediate catalyst is Google's intensified push into the AI chip market. Alphabet's Google is preparing to unveil a new generation of its Tensor Processing Units (TPUs) at the Google Cloud Next conference, with a specific focus on chips built for AI inference—the computational work of deploying trained models. Google Chief Scientist Jeff Dean emphasized the shift towards specialized hardware, stating, "It now becomes sensible to specialize chips more for training or more for inference workloads."
Google's ambition is backed by major customer agreements. Anthropic has signed a contract for 1 million TPUs, and Meta is using them through a multi-billion-dollar cloud agreement. Furthermore, Google is loosening access rules, allowing some customers to run TPUs in their own data centers and supporting external tools like PyTorch.
Beyond Google, billions in venture capital are flowing into AI chip startups, with the sector raising $8.3 billion globally in 2026. Notable fundraises include Cerebras ($1B), MatX, Ayar Labs, and Etched ($500M each). A key challenger is South Korean startup Rebellions, which just raised $400 million at a $2.34 billion valuation. Its Rebel100 chip is designed specifically for inference, and with investors like Samsung and SK Hynix, it has an edge in securing scarce high-bandwidth memory.
While Nvidia's GPUs remain the benchmark for AI training, and the company has moved to strengthen its inference offerings, investor confidence is being tested. Large hyperscaler customers like Microsoft, Alphabet, and Amazon continue heavy spending, but uncertainty persists around the sustainability of AI infrastructure investment and its near-term monetization. Despite the stock's underperformance relative to peers like AMD and Intel, Wall Street analysts, including Bernstein and Oppenheimer, maintain bullish long-term price targets for Nvidia.
