Nvidia nears $6t market cap as AI chip rally powers on

Nvidia shares rose as much as 4.7% to US$236.47 on May 14 as investors kept buying AI chip stocks.

The rally extended its seven-day gain to about 20% and pushed its market cap near US$6 trillion.

The Philadelphia Semiconductor Index has climbed nearly 70% since end-March as Intel and other chip stocks also rose.

Nvidia and Micron accounted for more than 30% of the S&P 500’s gain this year, adding to concerns that AI enthusiasm may be inflating a bubble.

Nvidia also drew attention after CEO Jensen Huang joined President Donald Trump on a China trip.

🔗 Source: Bloomberg

🧠 Food for thought

Implications, context, and why it matters.

Nvidia cuts many small chips from a semiconductor wafer. Cerebras Systems, an AI chip company, turns the full wafer into one giant chip 1.
That design packs in more computing capacity and uses on-chip SRAM (static random-access memory). Graphics processing units, or GPUs, usually rely on DRAM (dynamic random-access memory) 1.
Data moves across shorter paths on a single large processor. Cerebras says this can deliver AI responses up to 15 times faster than leading GPU-based systems 1.
The tradeoff is cost and complexity. That makes the design a niche option for demanding workloads, not a broad replacement for GPUs 1.

Attention is shifting away from the upfront cost of training AI models. The bigger issue now is the ongoing cost of running them for users 2.
That opens room for Cerebras in inference, the step where an AI system answers queries. Cerebras says its chips handle inference faster while using less power than Nvidia GPUs 2.
The same shift helps explain why Amazon and Google are building their own chips to take on part of the work now handled by Nvidia processors 2.
They want lower-cost computing options that use power more efficiently as they build the systems needed to keep AI models running and hold down expenses at scale 2.