On January 28, 2026, Representative John Moolenaar, chair of a key House committee, sent a letter claiming that Nvidia provided detailed technical guidance and software optimizations to China’s DeepSeek between 2023 and 2025. According to the document, Nvidia engineers offered algorithmic tuning that reduced DeepSeek’s training time by up to 40%, enabling the startup to develop its Pavo model, now reportedly deployed in PLA research facilities. The bipartisan concern centers on U.S. export controls for high-performance GPUs, with Moolenaar warning of an estimated $500 million in advanced chip shipments that may have directly supported military applications.

Nvidia will host its fourth-quarter and full fiscal 2026 earnings conference call on February 25, 2026 at 2 p.m. Pacific Time, with live listen-only webcast access for institutional investors. Ahead of the call, CFO Colette Kress will post detailed commentary on investor.nvidia.com immediately after the results are released at approximately 1:20 p.m. Pacific. Investors will be watching updates on data center revenue—previously up 66% year-over-year in Q3—and progress on the BioNeMo life sciences platform adoption by leading pharmaceutical partners, as well as the newly launched AI co-innovation lab with Eli Lilly and the Thermo Fisher autonomous lab collaboration.

Following a U.S. export license agreement requiring a 25% revenue share to the U.S. government, Chinese regulators have now green-lit domestic purchases of Nvidia’s H200 AI acceleration modules by major tech firms including ByteDance and Alibaba. Initial orders are expected to total several hundred thousand units, representing roughly $10 billion in revenue potential over the next 12 months. Chief Executive Jensen Huang has indicated in recent earnings commentary that Chinese sales could ultimately contribute up to $50 billion annually, making China one of Nvidia’s largest addressable markets once full logistics and customs processes are finalized.

Nvidia’s $20.6 billion acquisition of Groq in mid-2025 has generated a 230% return for Chamath Palihapitiya, who invested $62.3 million in the AI chip designer between 2017 and 2018. Despite the outsized gain, Palihapitiya admitted in a January interview that he struggled to celebrate, attributing his reaction to a focus on what might have yielded even higher multiples. Market analysts view Nvidia’s integration of Groq’s custom inference silicon as bolstering its competitive moat in data center AI workloads and expect the deal to contribute an incremental $1.2 billion to Nvidia’s data center segment in fiscal 2027.