About

World's leading AI chip company. H100 and GB200 GPUs power most AI training infrastructure globally.

Recent News

US lawmaker demands details on Trump's decision to allow Nvidia H200 chip sales to China

U.S. Rep. John Moolenaar (chair of the House China select committee) sent a letter to Commerce Secretary Howard Lutnick seeking the analysis behind the Trump administration’s decision to permit Nvidia to sell H200 AI chips to China. He argues that the policy shift risks eroding the U.S. advantage in aggregate AI compute—treating raw compute access as the true “engine of progress,” regardless of per-chip efficiency claims. The letter also points to reports that Huawei’s purported gains relied on chips allegedly obtained through intermediaries, framing that as a reason to tighten—not loosen—controls. The episode underscores how AI hardware export policy is becoming a fast-moving battleground where national security, industrial competitiveness, and supply constraints collide, and where each incremental allowance can reshape China’s near-term training capacity. ([reuters.com](https://www.reuters.com/world/us/us-lawmaker-demands-details-trumps-decision-sell-nvidia-h200-chips-china-2025-12-13/))

1d agoReuters

US stocks slide as Broadcom outlook rekindles fears of an AI investment bubble

US markets fell as investors reacted to signals that AI infrastructure growth may be running into margin pressure and capex reality checks, with Broadcom’s outlook amplifying concerns. The move dragged other AI-adjacent names and put the spotlight back on whether today’s spending pace can translate into durable profits, not just revenue growth. A key takeaway is that the AI trade is maturing: the market is starting to separate “AI demand exists” from “AI demand is profitable at scale,” especially for hardware/system sellers. For builders and buyers of AI, this kind of volatility tends to accelerate interest in efficiency—cheaper inference, better utilization, and more defensible unit economics.

1d agoReuters

Rivian rallies after detailing shift toward a custom autonomy chip and deeper AI integration

Rivian shares jumped after analysts responded positively to the company’s plan to move away from Nvidia processors for parts of its self-driving stack and to develop a custom autonomy/AI chip. The chip is set to be manufactured by TSMC and is positioned as a cornerstone for Rivian’s next vehicle platform (including the upcoming R2 line) and future “eyes-off” functionality timelines. This matters because autonomy is increasingly a compute-and-data advantage game: custom silicon can lower costs, improve latency, and let Rivian tune hardware to its model architecture. The risk, of course, is execution—custom chips are expensive, timelines slip easily, and safety validation remains the gating factor regardless of model capability.

1d agoReuters

Stock ImpactNVDA Price vs News

Deal Network

Loading deals...