Skip to content
16 November 2026

Jensen Huang resigns, JPMorgan flags rising competition for Nvidia

Nvidia’s founder departs, rivals close in, and a sprawling investment web adds new risk to the AI chip leader.

Jensen Huang resigns, JPMorgan flags rising competition for Nvidia

The board of Nvidia announced on Thursday that co-founder and chief executive Jensen Huang will leave the company after a series of personal scandals surfaced. The revelations include two extramarital affairs, one linked to a woman identified by U.S. authorities as a possible Chinese intelligence asset. Huang, who holds roughly 3.6% of the firm’s shares, will also vacate his seat on the board, prompting a rapid reshuffle of the leadership team.

Finance chief Colette Kress has been appointed interim chief executive while the board searches for a permanent successor. Former Synopsys head Harvey Jones will become the first chair of the board. Nvidia shares, which had dived more than 14% in early trading, recovered modestly as bargain hunters stepped in, but the company remains on track to lose close to $200 billion in market capitalisation. Rival chip makers AMD and Broadcom posted gains of roughly 6% each following the news.

Leadership turmoil and market reaction

The sudden exit of a CEO who was widely regarded as the architect of Nvidia’s AI surge creates uncertainty about execution rather than technology. Analysts note that the company’s product roadmap is largely public through 2028, but Huang’s personal role in coordinating a complex ecosystem of cloud partners, silicon suppliers, and research labs has been described as a competitive moat. Michael Feldman, portfolio manager at ClearBridge Investments, warned that losing that “conducting” capability could slow the pace of cash-flow generation even if the underlying chips remain market-leading.

President Donald Trump praised Huang’s contributions to U.S. technology leadership, even suggesting former Apple chief Tim Cook as a possible replacement. Lawmakers, including House Speaker Mike Johnson, emphasized the need for a swift and suitable successor given ongoing debates about tightening semiconductor exports to China. The episode also puts pressure on Commerce Secretary Howard Lutnick, who has faced criticism for not expanding the U.S. trade blacklist against Chinese firms.

JPMorgan flags growing competition from AI developers

In a separate analysis titled “The Icarus Files,” Michael Cembalest of JPMorgan highlighted a shift in the buying patterns of leading AI labs. Nvidia’s share of new computing contracts has slipped to about 39%, down from more than 80% during the 2023-2025 period, as developers increasingly explore in-house silicon and rival hardware. The report cites AMD’s recent gains in inference workloads, where memory bandwidth and inter-chip communication now rival Nvidia’s offerings, and Google’s Tensor Processing Units that performed competitively on China’s open-weight Qwen 3.5 model.

JPMorgan also underscored Nvidia’s exposure to Chinese open-weight AI models, which consume a substantial portion of global token volume. Proposed legislation such as the Remote Access Security Act (RASA) could force Asian cloud providers to obtain U.S. licences before granting Chinese customers access to restricted chips, potentially curbing a revenue stream that grew from 8% to 22% of sales with Singapore-addressed customers before the firm stopped reporting the metric in mid-2025.

Nvidia’s expansive investment and financing network

Beyond silicon sales, Nvidia has built a $124 billion ecosystem of equity stakes, acquisitions, and guaranteed financing that fuels demand for its products. Dealroom data show $99 billion of historic equity investments and $25 billion of planned commitments, ranging from $5 billion in Intel to multi-billion-dollar stakes in startups such as OpenAI, Anthropic, and Elon Musk’s xAI. The company also backs data-center infrastructure projects, including a $2 billion infusion into Lancium and similar pledges to SB Energy and Helix Digital Infrastructure.

These financial moves have been financed largely through operating cash flow, with Nvidia allocating roughly 58% of cash generated in the last three quarters to equity purchases and mergers. Off-balance-sheet debt and lease obligations surged from $214 billion to $422 billion between Q1 and Q2 2026, rising to $530 billion when land, power, and warranty guarantees are added. A change in this aggressive financing model could affect AI-compute providers such as CoreWeave and Nebius, while rivals like AMD and Broadcom might capture market share if Nvidia’s access to memory and logic wafers tightens.

Despite the turmoil, Nvidia still reports gross margins near 75% and operating margins around 66% for the 2023-2025 span, capturing 83% of AI-chip revenue. Its price-to-earnings-to-growth ratio sits at a historic 0.2, and the stock trades roughly 32% below its all-time high. The combination of a resilient balance sheet, long semiconductor design cycles, and a still-dominant CUDA ecosystem provides a cushion, yet investors must monitor leadership stability, competitive encroachment, and the far-reaching web of financing commitments.

Author

James Carter