Jensen Huang’s China Trip and Ineffable AI Deal

NVIDIA CEO Jensen Huang traveled to China this week at the personal request of U.S. President Donald Trump, joining a high-stakes diplomatic summit while simultaneously announcing a new AI partnership with British startup Ineffable Intelligence. The back-to-back developments — one geopolitical, one technical — place Huang at the center of two of the most consequential conversations in AI right now.

Trump Called Huang Directly to Join the China Summit

Huang’s inclusion in the U.S. delegation was not originally planned. According to CNBC, initial indications suggested Huang would not be part of the trip — but after media coverage highlighted his absence from the delegation, President Trump called Huang directly. Nvidia confirmed to CNBC that the invitation came from Trump personally.

Huang joined the delegation ahead of Trump’s scheduled meetings with Chinese President Xi Jinping on Thursday and Friday. Speaking to reporters on Thursday, Huang said Trump “asked me to come,” and described the summit as “one of the most important summits in human history,” according to a separate CNBC report.

The timing is significant. NVIDIA has faced sustained pressure from U.S. export controls that restrict sales of advanced chips — including the H100 and its successors — to Chinese buyers. Huang’s presence at a U.S.–China summit signals that semiconductor access remains a live negotiating issue between the two governments, even if no specific chip-trade terms were disclosed from the meetings.

Why NVIDIA’s China Exposure Matters

China has historically been one of NVIDIA’s largest markets for data center GPUs. Export restrictions introduced in 2022 and tightened in 2023 blocked sales of the H100 and A100 to Chinese customers without a license, pushing NVIDIA to develop downgraded variants — the H800 and A800 — specifically for that market. Subsequent rule changes in late 2023 restricted those chips as well.

The restrictions have created a measurable gap in NVIDIA’s addressable market. Chinese cloud providers and AI labs have responded by accelerating domestic chip development through companies like Huawei’s Ascend line and Cambricon, though analysts have consistently noted that domestic alternatives remain well behind NVIDIA’s performance benchmarks on most workloads.

Huang’s presence alongside Trump at the Xi summit does not guarantee any policy shift, but it signals that NVIDIA is actively engaged at the diplomatic level — not simply waiting on regulatory outcomes. No specific licensing changes or chip-sale agreements were reported from the summit as of publication.

Ineffable Intelligence Partnership: Reinforcement Learning at Scale

Separately, on Wednesday, May 13, NVIDIA announced a partnership with Ineffable Intelligence, a British AI startup focused on reinforcement learning — a training approach where models learn from experience rather than from static human-labeled datasets.

The distinction matters technically. Most frontier AI models, including large language models like GPT-4 and Claude, are trained primarily on text scraped from the internet and curated human feedback. Reinforcement learning, by contrast, allows a model to improve through iterative trial and error against an environment or reward signal — the same method that produced DeepMind’s AlphaGo and AlphaFold breakthroughs.

Huang framed the partnership in direct terms. “The next frontier of AI is superlearners — systems that learn continuously from experience,” he said in the announcement. CNBC reported that Ineffable is specifically focused on building AI systems that can learn continuously, rather than being trained once and deployed statically.

Financial terms of the partnership were not disclosed. It is not clear from the sources whether NVIDIA is taking an equity stake in Ineffable or providing compute access, engineering support, or some combination.

Intel and NVIDIA: A Ceremony, and a Tease

On a separate front, Intel CEO Lip-Bu Tan appeared alongside Huang at Carnegie Mellon University, where Huang delivered the commencement address and received an honorary doctorate. According to Wccftech, Tan — who performed the hooding ceremony — teased “exciting new products” developed in collaboration with NVIDIA.

No product names, specifications, or timelines were provided. The hint is nonetheless notable given the historically competitive relationship between the two companies in the data center chip market. Intel’s Gaudi accelerator line has struggled to gain traction against NVIDIA’s H100 and Blackwell-generation GPUs, and Tan has been working to rebuild Intel’s AI hardware credibility since taking the CEO role.

Carnegie Mellon University reported that Huang’s commencement address urged graduates to “shape what comes next” — a theme consistent with his broader public messaging around AI’s trajectory.

What This Means

Three threads are running simultaneously through NVIDIA’s week: geopolitics, a new technical bet, and a hint of cross-industry collaboration.

The China trip is the most immediately consequential. If the Trump–Xi summit produces any softening of export control restrictions — even for specific chip tiers — NVIDIA stands to benefit more directly than almost any other U.S. technology company. Huang’s personal inclusion, prompted by a presidential phone call, suggests the White House views NVIDIA’s market access as a legitimate bargaining chip in broader trade negotiations.

The Ineffable Intelligence deal reflects a longer-term strategic read. NVIDIA’s core business is selling compute, and reinforcement learning is among the most compute-intensive training paradigms — models trained this way require massive amounts of simulation and iteration. Backing a startup focused on continuous learning systems is a way for NVIDIA to seed demand for the next generation of GPU workloads, regardless of which specific model architecture wins.

The Intel tease is the least defined of the three, but worth watching. A genuine product collaboration between Intel and NVIDIA — historically fierce rivals — would represent a notable shift in the semiconductor industry’s competitive structure. Whether Tan’s comment reflects a real co-developed product or a loose partnership around interoperability remains to be seen.

FAQ

Why did Jensen Huang join Trump’s China trip?

According to CNBC, Trump called Huang personally after media coverage highlighted that the NVIDIA CEO was not included in the original delegation. Huang confirmed to reporters that Trump “asked me to come.”

What is Ineffable Intelligence and why is NVIDIA partnering with it?

Ineffable Intelligence is a British AI startup building systems based on reinforcement learning — where models improve through experience rather than static training data. NVIDIA’s partnership, announced May 13, positions the company to support what Huang called “superlearners” — AI systems that learn continuously, a workload that is particularly compute-intensive and thus well-suited to NVIDIA hardware.

What did Intel’s CEO say about a collaboration with NVIDIA?

At Carnegie Mellon University’s commencement ceremony, Intel CEO Lip-Bu Tan teased “exciting new products” developed with NVIDIA, according to Wccftech. No product names, specifications, or release timelines were disclosed.