NVIDIA CEO Transition: Who Is in Charge and What AI Customers Should Check

NVIDIA announced on November 12 that Jensen Huang has resigned as president, CEO and board member, effective immediately. CFO Colette Kress becomes interim CEO while retaining her finance role. The board says operations, customer commitments and the product roadmap should continue without interruption.

For an AI team buying hardware or renting compute, the immediate question is whether anything in its own deployment plan has changed. A leadership announcement does not answer that at the level of an order, a reserved cluster or a support escalation. Those need separate confirmation.

Who is running NVIDIA now?

According to NVIDIA's official transition announcement, Kress will serve until a permanent successor is named. The board is starting a search that includes internal and external candidates. Harvey C. Jones becomes chair of the board.

Kress joined NVIDIA in September 2013. Her company biography describes prior finance leadership at Cisco and Microsoft's Server and Tools division. She is an established member of the business, not an incoming executive learning its operating structure for the first time.

The appointment is interim. It should not be described as a completed permanent succession, and the announcement gives no date for choosing the permanent CEO.

What does the roadmap statement actually establish?

The board has publicly stated its expectation of continuity. That matters as NVIDIA's position, but it is not a new, product-by-product delivery schedule. The transition release does not provide revised hardware availability, software support dates or individual customer allocations.

A continuity statement is not an order confirmation. My recommendation is to keep existing technical plans grounded in their original evidence and check the few dependencies that could actually block delivery. Reworking an architecture because a company changed leadership would add cost without, by itself, addressing an identified engineering problem.

Equally, do not turn the board's statement into an unconditional promise to your customers. If your own launch depends on a future allocation, retain the distinction between a confirmed delivery and an assumption in the plan.

What to ask your supplier or cloud provider

Consider a team planning to move an inference service onto a reserved GPU cluster next month. Its useful questions are specific: is the reserved capacity still confirmed for the agreed date, is the quoted configuration unchanged, and who owns escalation if acceptance testing fails?

Ask the organization you contracted with. A cloud provider or systems supplier can clarify its own capacity and delivery commitment; an NVIDIA press release cannot confirm the terms of your agreement with that provider.

A short written check can cover the important dependencies without inviting a vague reassurance:

Please confirm whether our agreed delivery date, hardware configuration, support coverage and escalation contact remain unchanged. If any item has changed, identify the affected order or service and the revised commitment.

This is a suggested procurement message, not evidence that any customer's order has changed. Use it where a pending delivery is important enough to merit confirmation, not as a reason to interrupt every routine supplier relationship.

For a production service already running, start with your actual operating signals: available capacity, error rates, support responsiveness and upcoming renewals. Keep a dated record of a material change if one occurs. That gives engineering and procurement something concrete to act on instead of passing around headlines.

The next scheduled company update

NVIDIA's third-quarter results announcement schedules its fiscal 2027 Q3 release for approximately 1:20 p.m. Pacific on November 17, followed by a 2 p.m. call. The transition release says that schedule remains in place.

The public webcast is listen-only, with questions limited to financial analysts and institutional investors. Customers should not wait for that call to resolve an urgent order question. It is a place to look for a broader company update, not a substitute for the person responsible for your deployment.

The useful response to this transition is disciplined verification. Keep the leadership facts current, but make operational decisions against the commitments and measurements your application actually depends on.

I’m Paras. I work at the intersection of AI, engineering and distribution.

I help teams explain complex AI infrastructure and product changes clearly to technical and business audiences.

Book a 30-minute call.