What happened at the top
Huang resigned as president and chief executive of Nvidia, the company said on Thursday, November 12. A regulatory filing said the board accepted his resignation on Wednesday. CFO Colette Kress becomes interim CEO until a permanent successor is found, and she keeps the CFO job. The board is weighing both internal and external candidates.
Nvidia gave no reason for the departure. Outlets including ABC News reported it came days after the company announced an investigation tied to reporting that Huang was the target of a foreign intelligence effort, and after he admitted in an internal message that he had been unfaithful to his wife. This piece focuses on the business side. The personal and national-security questions are being covered extensively elsewhere.
For gamers, the relevant line came from board chair Harvey Jones. He said the company expects operations, customer commitments and product roadmap to continue without interruption. That is reassuring about upcoming products. It says nothing about pricing or how much priority GeForce gets.
The numbers behind the "afterthought" argument
The figures come from Nvidia's fiscal Q2 2027 report, as summarized in the research behind this piece. I could not open the filing directly, so treat these as reported rather than checked line by line.
- Total revenue was $96.221 billion.
- Data Center revenue was $89.023 billion, which works out to about 92.5 percent of the total.
- Data Center growth was attributed to Blackwell Ultra infrastructure.
- Edge Computing revenue was $7.198 billion. Nvidia said strong Blackwell workstation sales were partly offset by slower consumer PC sales, which it blamed on elevated memory and systems prices.
- Nvidia guided fiscal Q3 revenue to about $108 billion, assuming no China data-center compute revenue. That is a forecast, not a result.
Two caveats matter here:
- 92.5 percent is not "AI versus gaming." It is Data Center's share of total revenue. The other 7.5 percent is not all gaming, because Nvidia splits its revenue across several market platforms.
- The gaming share is an estimate. PCWorld cites analyst Ian Cutress of More Than Moore, who puts gaming at roughly 5 to 7 percent of revenue, possibly less. That is his estimate, not an Nvidia disclosure.
From GeForce to CUDA
Nvidia's own timeline says it was founded on April 5, 1993, by Jensen Huang, Chris Malachowsky and Curtis Priem. The original aim was to bring 3D graphics to gaming and multimedia. The company says it invented the GPU in 1999, which PCWorld ties to the GeForce 256. Nvidia credits CUDA in 2006 with opening GPU parallel processing to general-purpose computing, and that is the bridge to today's AI business.
Gaming was the foundation, and the same parallel-processing approach turned out to suit far more than rendering. Nvidia says that in its filings it now describes itself as a data center-scale AI infrastructure company.
Does AI demand really starve GeForce?
Chacos's column makes a careful version of this argument, and the limits of the evidence matter.
What the evidence supports:
- Data centers are not buying GeForce cards off store shelves. Chacos says so directly.
- AI accelerators and gaming cards use different memory and packaging. AI parts typically use HBM and advanced packaging, while GeForce uses GDDR. They are separate production lines that still compete for some upstream resources and investment.
- Cutress says memory vendors talk almost entirely about HBM and DDR or LPDDR, not GDDR. He also says GDDR8 isn't yet standardized and is expected toward the end of the decade.
- Nvidia's filing reports supply and capacity commitments rising from $119 billion to $279 billion as of July 26, 2026. That shows how heavily the company has committed to data-center demand.
- Chacos notes that Nvidia announced expansions of AI-oriented packaging, substrates and testing capacity last month. He cites a $1.1 billion advance for new testing facilities.
What it does not prove:
- The column includes a line saying RTX 5090 dies are going to AI data centers. That is a rhetorical flourish. The sources reviewed do not establish that specific diversion.
- The crypto comparison is an analogy about who wins when buyers with more money compete for graphics hardware. The mechanism today is different.
Nvidia's financial filings alone cannot show a direct trade-off between AI customers and gamers. The incentive is plain, though. Every wafer, memory allocation and engineering hour has an opportunity cost, and the company's growth is overwhelmingly on the data-center side.
Will a new CEO change GeForce's fortunes?
Cutress told Chacos that gaming persists at Nvidia partly because it is tied to the company's legacy, and that legacy is tied to Huang, who is a gamer himself. He added that some investors would rather see GeForce gone entirely. He thinks an Nvidia veteran such as CUDA co-creator Ian Buck would be more likely to keep gaming going than an outsider brought in to squeeze the business. Yahoo Finance reported Buck emerging as a frontrunner on prediction markets. That is market speculation, not an announcement.
Technology also flows both ways. Cutress points to DLSS as an example of gaming features benefiting from Nvidia's AI research, which gives the company a practical reason to keep GeForce going. The counterpoint is price. Chacos calls the RTX 5090's $1,999 launch price a luxury tax on owning Nvidia's best silicon.
There is also a near-term test. Technology.org reports Nvidia is due to report quarterly results on November 17, days after the leadership change. That report and any commentary on the consumer side will show how much interim leadership signals about gaming.
Bottom line
- Nvidia's quarterly results show data centers at about 92.5 percent of revenue, so the company's center of gravity is no longer gaming.
- Gaming's exact share is not disclosed in the cited results. The 5 to 7 percent figure is an analyst's estimate.
- A change at the top does not remove the financial incentives favoring AI hardware.
- Nvidia says its roadmap continues uninterrupted. Whether that includes better-value GeForce cards is the open question.
For PC builders, the practical takeaway is to watch memory pricing and GDDR supply as closely as GPU announcements. Per Cutress, those are where the AI boom's effect on gaming hardware is most visible.