Nvidia to Ship 40,000–80,000 H200 GPUs to China, Unlocking $1.3–2.6 Billion Q1 Revenue
Reuters reports Nvidia will resume H200 GPU shipments to China by mid-February, shipping 40,000–80,000 units valued at $1.28–2.56 billion in Q1 fiscal 2027. Under a U.S. agreement, Nvidia will remit 25% of Chinese sales and plans to bring new capacity online and take orders in Q2 2026.
1. Groq Partnership Removes Final Bear Case
Bernstein analyst Stacy Rasgon argues that Nvidia’s recently announced non-exclusive licensing agreement with AI-startup Groq eliminates the last significant bearish argument against NVDA shares. Groq, founded by Jonathan Ross—the architect of Google’s first tensor processing unit—specializes in high-bandwidth, low-latency inference architectures. Under the pact, Nvidia will integrate Groq’s line processing units into its AI ecosystem, reinforcing its leadership in both training and inference. Industry sources estimate the deal could be worth up to $20 billion over its multi-year term, underscoring Nvidia’s ability to preempt emerging competitors and solidify its moat in the rapidly evolving AI chip market.
2. Chinese H200 Shipments Signal Revenue Rebound
A Reuters report indicates that Nvidia plans to resume shipments of its advanced H200 data-center GPUs to approved Chinese customers beginning before the Lunar New Year holiday in mid-February. Initial volumes are expected between 40,000 and 80,000 units, representing $1.28 billion to $2.56 billion in first-quarter fiscal 2027 revenue based on prevailing list prices of approximately $32,000 per processor. Nvidia will allocate existing H200 inventory to fulfill these orders and ramp additional capacity in Q2 2026. This reopening of the Chinese market follows export restrictions in April 2025 and could add materially to Nvidia’s already record-high data-center backlog, estimated at $275 billion.
3. Recent Pullback Offers Attractive Entry Point
After briefly reaching a $5 trillion market capitalization at October’s end, Nvidia shares have retraced roughly 10%, bringing its valuation closer to long-term growth expectations. The stock now implies a forward earnings multiple in the mid-20s, compared with historically higher levels. Nvidia’s gross margin remains industry-leading at just over 70%, and its vertically integrated product suite continues to lock in hyperscale cloud providers and enterprise customers. The pullback has prompted some strategists to upgrade their ratings, highlighting the combination of sustained demand for AI infrastructure and a more tempered valuation as ideal conditions for new investors.
4. Billionaire David Tepper’s NVDA Accumulation
Hedge fund manager David Tepper significantly expanded his Nvidia position in 2025, increasing holdings from 300,000 shares at the end of Q1 to 1.9 million shares by Q3, according to SEC Form 13F filings. This more than sixfold increase makes Nvidia one of the fund’s top four equity positions. Tepper’s move underscores confidence in Nvidia’s ability to meet surging demand for GPUs amid an extended AI buildout. With global data-center capital expenditures projected to rise to $3–4 trillion annually by 2030, Tepper’s aggressive accumulation suggests he views the recent price dip as a strategic opportunity to capitalize on long-term secular growth in AI computing.