Nvidia’s Quiet Exterior Masks a Two-Pronged Pivot: A China CPU and a Healthcare AI Bet

Nvidia shares hover near 50-day moving average while it pursues a CPU-based China strategy via the Vera chip and deepens healthcare AI with Abridge.

Nvidia’s stock ended the week at €177.28, virtually unchanged and sitting almost precisely on its 50-day moving average of €177.30. On the surface, the tape suggests a pause. Underneath, the company is orchestrating two distinct strategic moves — one aimed at reclaiming the Chinese market, the other at embedding itself deeper into the healthcare industry.

Vera CPU: A regulator’s blind spot

After seeing its GPU market share in China effectively collapse — CEO Jensen Huang admitted in October that it had fallen to zero — Nvidia is now trying a different route. Instead of graphics processors, it is offering a central processing unit. The company has informed Chinese clients that its new Arm-based “Vera” chip for AI data centres could be available from August, and it is already taking orders.

The logic rests on a regulatory gap. Washington’s export restrictions target GPUs; CPUs fall into a less restricted category. While around ten Chinese companies have been cleared to buy the H200 GPU, not a single unit has been delivered because Chinese authorities are blocking approvals. Vera offers a workaround by competing directly with Intel and AMD in a less politically exposed product class.

Initial demand is tangible but modest. One large Chinese cloud provider is preparing an order for more than 300 dual-Vera servers, though it plans to test them only in data centres outside China first. Software and migration hurdles are slowing broader adoption. Nvidia has set a revenue target of $20 billion for Vera by the end of its fiscal year in January 2027.

Abridge: A seat at the doctor’s table

On June 11, Nvidia announced a partnership with Abridge, a health-tech company valued at $5.3 billion that counts more than 300 health systems as clients, including Kaiser Permanente, Johns Hopkins Medicine and Yale New Haven Health. The collaboration will build a clinical AI model on Nvidia’s open Nemotron architecture.

Abridge’s app listens to patient-doctor conversations and automatically converts them into clinical notes. The new model aims to improve that documentation and support clinical decision-making. Nvidia is no newcomer to this relationship — its venture arm has participated in several of Abridge’s funding rounds. The partnership deepens an existing tie while Abridge simultaneously announced other deals, including one with Eli Lilly.

Healthcare represents more than a side project. Nvidia has long been the infrastructure supplier for AI’s boom, capturing a large slice of every dollar hyperscalers pour into GPU clusters. But investors are increasingly questioning how long those capital expenditures can keep rising, whether Amazon, Google and Microsoft will erode Nvidia’s market share with custom chips, and what happens if the Chinese market stays shut. Healthcare offers a separate revenue stream built on Nvidia’s open-model ecosystem rather than on silicon alone. Microsoft already works with the Mayo Clinic on a medical AI; OpenAI and Anthropic have launched health products. Nvidia is now securing its own spot in the examination room.

The product cycle and the financial backbone

Meanwhile, the next hardware generation is rolling out. Vera Rubin, the successor to the Blackwell architecture, is in full production. First Rubin-based products should reach partners — among them AWS, Google Cloud, Microsoft, Oracle Cloud, CoreWeave, Lambda and Nebius — in the second half of 2026. Nvidia claims Rubin will be five times faster than Blackwell on inference and 3.5 times faster on training. On the last earnings call, management forecast that hyperscaler AI capex will hit $1 trillion by 2027.

The financials remain staggering. In the first quarter of fiscal 2027, Nvidia posted a record $81.6 billion in revenue, up 85% year over year. The board raised the quarterly dividend from $0.01 to $0.25 per share — a twenty-five-fold increase — and authorised an additional $80 billion in share buybacks. The dividend will be paid on June 26 to shareholders of record on June 4. In the past quarter alone, the company returned roughly $20 billion to shareholders through buybacks and dividends combined.

What the price says — and what it doesn’t

The stock’s stagnation at €177.28, almost exactly on its 50-day average, masks the underlying tension. The relative strength index sits at 45.6, signalling neither euphoria nor panic. The all-time high of €202.50, set in mid-May, is about 12% away. Nvidia can deliver the largest quarter in technology history, yet the share price gives ground because expectations are already priced for perfection.

Analyst sentiment remains strongly positive. A consensus of 62 analysts gives the stock a “Strong Buy” rating and a target of $298.42. Another measure puts the average price target at €258.25, implying roughly 46% upside from current levels. The gap between those forecasts and the market price reflects how differently the same data can be interpreted.

The final weeks of June shape up as a dense calendar for Nvidia shareholders: the annual general meeting on June 24, the dividend payment on June 26, and the possibility of initial Vera deliveries to China. The twenty-five-fold dividend hike is not a routine adjustment — it is a statement of confidence in the company’s cash flow, backed by a new CPU that tries to reopen the world’s second-largest economy and a healthcare partnership that takes Nvidia into the clinic.