Nvidia Backs Marvell-Lumentum Tie-Up While Tightening the Screws on Cloud Partners' China Exposure

Nvidia has thrown its weight behind a major consolidation within its own supplier base — and, on a separate front, is pressing the cloud providers that rent out its chips to lean away from Chinese customers. The two developments, reported within a day of each other, sketch a company working both sides of its supply chain: shaping who builds the hardware and who gets to use it.

On the deal side, Bloomberg reported that Marvell CEO Matthew Murphy and Lumentum CEO Michael Hurlston consulted Nvidia chief Jensen Huang by phone before unveiling their combination. Hurlston described the response as positive. Nvidia's interest is not purely advisory: the chipmaker put $2 billion each into Marvell and Lumentum in March 2026, making it both an investor in and a customer of the two companies.

The all-stock transaction, announced a day earlier, carries an enterprise value of roughly $115 billion including debt. It would fold Marvell's custom silicon together with Lumentum's optical components, giving Nvidia what it hopes will be a broader product range from a single source. Closing is targeted for the first half of 2028, subject to shareholder and regulatory approvals.

That timeline matters. Huang's endorsement signals support for the plan, not confirmation of operational gains already in hand. Whether the combined company can actually deliver faster and wider is a goal, not a demonstrated result — and no specific revenue or profit contribution for Nvidia can be quantified from the announcement. Investors, for their part, gave the news a chilly reception: both Marvell and Lumentum fell sharply the previous day, a stark contrast to Huang's warm assessment.

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

A 51% Threshold for Future Contracts

The second thread runs through Nvidia's commercial relationships. According to the Financial Times, cited by Reuters — which said it could not independently verify the details — Nvidia has spent several months verbally instructing cloud partners worldwide to reduce their reliance on Chinese customers. Under the guidance, at least 51% of services in new contracts must go to businesses outside China, capping the permissible Chinese share at 49%.

Five people familiar with the matter are the FT's source. The rules are said to apply globally but only to new agreements; existing computing-capacity leases remain untouched. That distinction is central — the report describes no immediate withdrawal from all dealings with Chinese clients, but rather a tighter grip on the customer mix partners can sign up going forward.

Execution falls to Nvidia's sales and regional teams. For affected cloud providers, the guidance adds a layer of commercial friction: winning demand for compute is no longer enough, since new deals must also pass a test on where the customer is based. Providers in Singapore have warned, per the report, that finding enough non-Chinese customers will be difficult, and strict enforcement could dent their business. The risk for Nvidia therefore extends to partners outside China whose demand ultimately traces back to Chinese users.

Controls That Reach Past the Chip Sale

The reported requirements sit within a wider tightening of US restrictions on advanced AI technology and a crackdown on smuggling networks. The Commerce Department issued related guidance in May, with Blackwell chips among the products covered. Nvidia had already cut its list of approved Asian customers by more than 50% in July, and its vetting includes data-center visits, contract checks and conversations with end users.

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

What emerges is a compliance regime that follows the silicon well beyond the point of sale — down to how cloud providers' own customers use the compute. That could stretch the commercial reach of export rules further than individual chip shipments would suggest. For investors, the key variable is scope: existing leases appear protected under the report, while future agreements face heavier constraints. Concrete revenue effects cannot yet be tallied, and Reuters' inability to confirm the details independently remains an open caveat.

Read together, the two stories point to the same underlying question for Nvidia shareholders — how the company's investments and its supply relationships translate into returns. Backing a merger among key suppliers is one lever; dictating the customer base of the clouds that deploy its hardware is another. Neither, so far, comes with a number attached.

Ad

Nvidia Stock: New Analysis - 3 November

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...