Home » Tech News November 13, 2026: Nvidia Loses Jensen Huang, Gemini 4 Hydrogen Arrives, and Zelda Breaks Records
Posted in

Tech News November 13, 2026: Nvidia Loses Jensen Huang, Gemini 4 Hydrogen Arrives, and Zelda Breaks Records

Futuristic AI data center with glowing GPU chips and a city skyline at dusk, symbolizing a day of big changes in artificial intelligence

Tech News November 13, 2026: Nvidia Loses Jensen Huang, Gemini 4 Hydrogen Arrives, and Zelda Breaks Records

Thursday was a day when the tech world changed all at once — and the headline is that the Nvidia CEO resigns after 33 years at the top. Google shipped a faster coding-agent model, SpaceX sketched out more than $74 billion in AI campuses, AMD took a swing at Nvidia’s newest chips, and a 28-year-old Zelda game set a sales record.

Key Takeaways

  • Nvidia CEO Jensen Huang resigned; CFO Colette Kress is interim CEO.
  • Google’s Gemini 4 Hydrogen: a faster, cheaper model for coding agents.
  • SpaceX plans $74 billion-plus in AI campuses in Texas and Arizona.
  • AMD claims its MI455X beats Nvidia’s Vera Rubin — but AMD ran the test.
  • Zelda: Ocarina of Time sold 5.2 million copies in four days — a Switch 2 record.

Nvidia CEO Resigns: Jensen Huang Steps Down After 33 Years

Nvidia announced Thursday that Jensen Huang has resigned as president, chief executive and board member, effective immediately; the board accepted his resignation the day before. Huang co-founded Nvidia in 1993 and led it for 33 years, turning a graphics-chip startup into the heart of the AI boom.

According to Investopedia, CFO Colette Kress will serve as interim CEO while keeping her finance job, and the board will consider internal and external candidates. Veteran director Harvey C. Jones becomes Nvidia’s first-ever board chair. Shares fell on Thursday as investors digested the news.

Nvidia reports third-quarter results Tuesday, November 17, and Kress will front that call. The resignation followed reports about Huang’s personal conduct and a board investigation opened November 8; the announcement gave no reason. For context, see our November 9 roundup.

Google Launches Gemini 4 Hydrogen for Coding and AI Agents

Google announced Gemini 4 Hydrogen on November 12, a faster, cheaper member of the Gemini 4 family built for coding and agentic work. According to Google, it rolls out across Google AI Studio, the Gemini API, Search and other Google products at $1.50 per million input tokens and $7.50 per million output tokens.

Google also launched a Gemini Decisions API, with a Decision 1 model for fast yes/no and multiple-choice calls at $0.10 per million input tokens.

As agents take on more real work, their models need to be fast and cheap, not just smart. Hydrogen is Google’s bet on being that workhorse.

SpaceX Plans $74 Billion in AI Supercomputing Campuses

SpaceX plans to invest more than $74 billion in two AI supercomputing campuses, Reuters reported: over $24 billion at its Starbase site in Texas and about $50 billion in Cochise County, Arizona.

The Arizona campus still needs county approval, and would create around 5,500 construction jobs and 1,000 permanent ones. The Texas facility would generate its own electricity instead of drawing from the public grid.

These are proposed investments, not finished buildings — but the scale is the story.

AMD’s MI450X Takes Aim at Nvidia’s Vera Rubin

AMD announced a new Instinct MI450X accelerator on November 12, with 288 GB of HBM4 memory, plus fresh inference results for its MI455X chip in the Helios rack system. According to Engineering.com, AMD’s labs measured 21,333 tokens per second per GPU on DeepSeek-R1 — up to 30% higher than Nvidia’s published Vera Rubin NVL72 result.

One caveat: AMD ran the test itself. Same benchmark family, same model — but AMD’s own result against a published Nvidia number, not an independent head-to-head run.

Even so, the MI450X gives Helios customers a second memory option — and the timing, on the day Nvidia’s CEO resigned, was hard to miss.

Anthropic Lets Companies Run Claude Agents on Their Own Servers

Anthropic opened a private beta on November 12 for self-hosted Claude Managed Agents: companies can now run the agent service — harness, sandboxes, credentials and session history — on their own infrastructure. According to Anthropic, the model itself stays put: every call still goes to Anthropic, Amazon Bedrock, Google Cloud, Microsoft Foundry or the customer’s own gateway, and Anthropic says it has no remote access to the self-hosted server or its logs.

Access is by request through an Anthropic account team; no pricing or general-availability date yet. For banks, hospitals and governments, it removes a big blocker — agent history no longer has to live on a vendor’s servers.

Zelda’s Ocarina of Time Remake Sets a Switch 2 Sales Record

Nintendo announced that The Legend of Zelda: Ocarina of Time sold more than 5.2 million copies worldwide in four days after its November 5 launch. According to Nintendo, it is the fastest-selling Nintendo-published Switch 2 game globally and in the U.S., Canada and Latin America, including 810,000 in Japan.

It is already the Switch 2’s second-best-selling game, behind Mario Kart World’s 16.48 million — passing Donkey Kong Bananza’s 5.05 million lifetime total in under a week.

The 1998 original is one of the most beloved games ever made, and this remake rebuilds it with overhauled visuals. Nostalgia is a hell of a launch strategy.

The Takeaway

Three threads ran through Thursday: leadership at the top of the AI stack is suddenly uncertain, the infrastructure arms race keeps escalating, and AI keeps getting cheaper, faster and more self-hosted. And a 28-year-old Zelda game outsold nearly everything new.

Frequently Asked Questions

Why did Jensen Huang resign as Nvidia CEO?

Nvidia announced on November 12 that Huang resigned as president, CEO and board member, effective immediately; the board had accepted his resignation on November 11. The announcement gave no reason. It followed reports about his personal conduct and an independent board investigation opened on November 8. CFO Colette Kress is interim CEO.

What is Google Gemini 4 Hydrogen?

A faster, lower-cost model in the Gemini 4 family, built for coding and AI agents. It is available through Google AI Studio, the Gemini API and Search, priced at $1.50 per million input tokens and $7.50 per million output tokens.

How much is SpaceX investing in AI campuses?

More than $74 billion: over $24 billion at its Starbase site in Texas and about $50 billion in Cochise County, Arizona. The Arizona campus still needs county approval.

How many copies did the Zelda Ocarina of Time remake sell?

More than 5.2 million worldwide in its first four days, making it the fastest-selling Nintendo-published Switch 2 game.

Sources

Leave a Reply

Your email address will not be published. Required fields are marked *