Nvidia’s Waiting Game: Vera Rubin Looms as China Stays Locked Out

Nvidia’s stock has settled into a familiar holding pattern, hovering around €183 — roughly 10% below the all-time high of €202.50 hit in May. The dip isn’t a sign of crumbling fundamentals. It’s the market catching its breath, waiting on the most consequential product cycle in the company’s history while a once-lucrative market remains stubbornly off-limits.

Over the past twelve months, the shares have rallied about 45%. Since January, they’ve added 13%. The average analyst price target of €257.45 implies further upside of more than 40%, yet the stock’s relative strength index of around 50 signals a market in equilibrium — neither euphoric nor panicked. The annualized 30-day volatility of 42% underscores how sharply every geopolitical headline feeds into the price.

The Rubin Catalyst Takes Shape

The reason for the quiet confidence has a name: Vera Rubin. Nvidia has confirmed full production of the next-generation AI platform, with initial partner deliveries due in the second half of 2026. The performance leap is stark: Rubin delivers five times the inference throughput of current Blackwell systems while slashing cost per token by a factor of ten. On the GTC 2026 stage, Jensen Huang declared this the “Year of Inference,” marking a shift from training mega-models to real-time deployment. The first Vera Rubin rack is already running at Microsoft Azure, though the real volume ramp is expected only toward early 2027.

That timeline explains why the stock has eased nearly 5% over the past month. Patient capital is waiting for shipment confirmations. No one is throwing the core thesis overboard. The numbers back it up: Nvidia posted a record $81.6 billion in revenue for the first quarter of fiscal 2027, up 85% year-on-year, with the data center segment soaring to $75.2 billion on the back of Blackwell chips. About half of that came from the big cloud providers, while the rest is spreading across an ever-widening customer base.

Should investors sell immediately? Or is it worth buying Nvidia?

Sovereign AI: A Geopolitical Demand Driver

A new demand pattern is also emerging. According to the Center for a New American Security, Nvidia now powers 52% of all government AI infrastructure projects worldwide. Sovereign AI — national governments building their own data centers and locally controlled systems — is pouring billions of dollars into Nvidia’s order book. This demand is structurally different from hyperscaler spending: it follows geopolitical imperatives, not quarterly return calculations, making it far more durable.

The China Dead End

But for all the cheer in the data center, the China question remains unresolved — and it’s not small. Washington has granted roughly ten Chinese companies, including Alibaba, Tencent, ByteDance, and JD.com, licenses to buy Nvidia’s H200 chips, with each approved buyer allowed up to 75,000 units. Lenovo and Foxconn are authorized as distributors.

It sounds like a market opening. It isn’t.

Nvidia’s CFO stated bluntly on an earnings call that while the company has received permission for small H200 volumes to Chinese customers, it has not booked a single dollar of revenue from those approvals. The decisive line: “We do not know whether any imports will be allowed into China.” The bottleneck lies in Beijing’s strategic hesitation. A Trump-era agreement reportedly stipulates that 25% of chip revenue would flow back to the U.S., a step China views as a security risk — creating extra exposure along the supply chain.

Meanwhile, China’s AI infrastructure budget continues flowing to Huawei and domestic champions. Cities like Hangzhou offer subsidies to companies running AI models on local chips, financially cementing the incentive to avoid American technology. Jensen Huang acknowledged in April 2026 that Nvidia’s market share in China has effectively dropped to zero — a stunning fall from the roughly 95% it once commanded for advanced AI accelerators. As recently as fiscal 2025, China contributed about 13% of total revenue.

Huang accompanied President Trump on his May 2026 state visit to China, visibly hoping for a breakthrough on H200 shipments. Goodwill was generated. Confirmed deliveries? None.

Nvidia at a turning point? This analysis reveals what investors need to know now.

Hyperscalers: Frenemy or Foe?

A separate pressure point comes from Nvidia’s biggest customers. Amazon, Google, and Microsoft are all developing their own custom silicon — TPUs, Trainium, and the like. Their combined share of the accelerator market is expected to rise from roughly 21% in 2025 to nearly 28% in 2026. That could cap Nvidia’s growth among the hyperscalers. Yet those same companies remain Nvidia’s largest buyers. This structural tension has so far consistently resolved in Nvidia’s favor, and the company still commands an estimated 85–92% of the AI accelerator market. AMD and Intel together account for single-digit shares in the training segment.

Two Uncertainties, One Stock

The stock is now trading just 3% above its 50-day moving average of €177.93. The RSI of 51.5 reflects a market that is watching, not rushing. The analyst target of €257.45 likely doesn’t fully price in the China risk — but neither does it capture the optionality. A market that once generated 13% of revenue, is now virtually zero, and could theoretically reopen at any moment. That optionality is unbaked, unresolved, and perhaps the most intriguing part of the Nvidia story right now.

The coming months will deliver clarity on both fronts: the Rubin ramp in the second half of 2026 and any movement on China’s import stance. Until then, Nvidia’s investors are content to wait.

Ad

Nvidia Stock: New Analysis - 16 June

Fresh Nvidia information released. What's the impact for investors? Our latest independent report examines recent figures and market trends.

Read our updated Nvidia analysis...

en | US67066G1040 | NVIDIA’S | boerse | 69551563 |