Nvidia Establishes National Security Committee as Rivals Seize Market Share
nashnova research
Nvidia established a board-level national security committee in its first major post-Jensen Huang move, responding directly to congressional scrutiny; meanwhile AMD and Google are accelerating their push into Nvidia's AI chip territory, with every day of leadership vacuum reshaping the competitive landscape.
What does this committee actually do?
Nvidia announced a standing National Security and Export Control Committee, chaired by director Suzanne Nora Johnson, meeting at least once per quarter to oversee export compliance and global trade risk at board level.
This means → Nvidia elevated export controls from a management-level task to a board-level institutional mechanism, signaling to Congress and regulators that it is treating the issue seriously.
The backdrop: Congress is investigating Huang's ties to alleged Chinese intelligence figures, and Huang has agreed to testify before Congress this month.
Why did the new leader's first move target compliance, not business?
Interim CEO Colette Kress — Nvidia's former CFO — made the committee her first major public action after formally taking the helm, timing it hours before the fiscal Q3 earnings release.
In plain terms = Kress chose to put out the fire first — using an institutional move to ease congressional and market concerns about compliance, then letting the earnings numbers stabilize investors.
On the same day, the U.S. Commerce Department issued an interim rule requiring data-center operators in Singapore, Malaysia, Thailand, Indonesia, and Brazil to obtain licenses before purchasing certain chips, widening the export-control net further.
How are AMD and Google seizing the moment?
According to The Information, AMD's penetration of the AI inference market began well before Huang's departure; the leadership upheaval is accelerating the trend. Superlinked CEO Daniel Svonava called Kress's appointment "the last straw" that pushed his firm from testing AMD to actual deployment.
AMD's edge runs along two lines: lower inference costs than Nvidia, especially attractive for vendors running open-source small models; and a markedly improved software ecosystem — key cloud provider TensorWave says staff time spent helping clients debug RocM has dropped 80–90% year-on-year.
Google is also expanding external sales of its tensor processing units (TPUs — custom AI chips Google designs in-house), landing OpenAI and Oracle as clients and taking on up to $44 billion in data-center lease-guarantee exposure to do so.
Can the earnings numbers hold the line?
Wall Street expected Nvidia's fiscal Q3 data-center revenue to nearly double to roughly $101.87 billion, with total quarterly revenue guidance of about $125.2 billion.
But the risk exposure is explicit: UBS analysts flagged that in a worst case, Asia shipping controls could erode roughly 10% of revenue; HSBC downgraded Nvidia from "buy" to "hold" last week, noting that regions outside the U.S. and Taiwan account for 11% of total revenue — all directly exposed to export restrictions.
This means → the numbers themselves may look strong, but what the market is really watching is Kress's commentary on the earnings call about export-control exposure, Asian order allocation, and willingness to deploy capital — the first real test of how "post-Huang" Nvidia gets repriced.
Could Jensen Huang still come back?
According to the Financial Times, President Trump said before departing for an Asia summit that he "wouldn't mind" Huang returning; the FT also reported that Huang had discussed a possible comeback with close allies.
But people familiar with the matter said that as of Monday afternoon, reinstating Huang as CEO has been explicitly ruled out.
Nvidia's stock has fallen roughly 16% since the scandal broke on November 8. This reflects a market that is repricing the company — shifting from "the Jensen premium" toward a reassessment of institutional governance, compliance, and the competitive landscape.
市场有风险,内容仅供研究参考,不构成投资建议。
