
During the prediction segment of Prof G Markets, host Ed Elson asked whether Jensen Huang would return to Nvidia. Rohan Goswami put the probability at 0%. His assessment rested on two sets of verifiable facts. First, he had recently attended an industry event bringing together around 400 AI and technology company CEOs, where the possibility of Huang’s return was a major topic of discussion. Second, according to his earlier reporting, Nvidia’s board deliberated for hours before deciding on the leadership change. Some directors had worked with Huang for more than 30 years, with relationships stretching back to the company’s founding. Goswami’s view was that, having gone this far, the board would be unlikely to reverse course simply because of outside pressure.

Deirdre Bosa, who appeared on the same programme, offered a different perspective. She noted that Huang had resigned voluntarily rather than being directly dismissed by the board, potentially leaving room for a future return. Sam Altman’s return just days after being removed as OpenAI’s CEO also led some observers to believe a similar scenario could unfold. However, she acknowledged that public support—including Elon Musk signing a petition calling for Huang’s reinstatement and Donald Trump saying his return “might not be a bad idea”—did not necessarily mean the board would change its decision.
The suggestion that Huang could launch another venture with $50 billion was Goswami’s personal prediction on the programme, rather than a plan Huang had announced. But the figure was not arbitrary: it matched the size of Nvidia’s $50 billion AI financing platform, which Huang personally unveiled in August 2026.
Nvidia Signals There Is No Going Back
Nvidia Chairman Harvey Jones explicitly stated on a conference call that the board had no intention of bringing Huang back. He said the company was selecting a long-term CEO and had appointed executive search firm Heidrick & Struggles to lead the process. This was more than a vague statement that “all options remain open.” It was a clear position backed by concrete action: an external search firm had been engaged, and the selection process was underway.
This marks a crucial difference from the Altman precedent Bosa mentioned. Altman returned to OpenAI within days of being dismissed, with his reinstatement resulting from a reversal of the board’s earlier decision. Nvidia’s situation is different: Huang resigned voluntarily, after which the board launched a formal search and publicly ruled out his return. When a company’s chairman tells analysts on an earnings call that there is “no intention of bringing Huang back,” the statement serves more than simply answering speculation. It also reassures customers, employees and executive recruiters that the objective is to find a new CEO, rather than wait for the founder to return.
Goswami’s assessment—that the board would be unlikely to reverse course under outside pressure after going this far—was reinforced by these formal steps. Musk’s petition and Trump’s verbal support cannot change the fact that an executive search is already underway.
Speculation About Huang’s Next Move
Goswami’s decision to link Huang’s next move to the $50 billion figure rests less on Huang’s personal fortune—an estimated net worth of $145 billion—than on what happens to Nvidia’s $50 billion AI financing platform following its founder’s departure.
Huang personally unveiled the platform in August 2026. Its goal was to pool capital from multiple sources to provide $50 billion in financing for AI computing deals. The platform’s central asset was not the money itself, but Huang’s personal credibility and network as its chief dealmaker. Whether participants were willing to finance data centres and chip assets through the platform depended heavily on their trust in his judgement.
Market institutions have already highlighted the uncertainty facing the platform after Huang’s departure. Reuters reported that losing its chief dealmaker could make it harder for Nvidia to finance some of its riskier infrastructure bets. UBS analyst Tim Arcuri offered a more specific estimate: if Nvidia tightened its financing support, $50 billion or more in funding could disappear from the market.
This means the $50 billion in Goswami’s prediction is on the same scale as the potential funding gap facing Nvidia’s financing platform. If Huang were to launch a new company with $50 billion, he could be drawing on a capital network migrating from the old platform, rather than raising funds entirely from scratch. AI infrastructure projects that previously obtained financing through Nvidia’s platform, along with investors who entered the ecosystem because they trusted Huang’s judgement, would face a choice: follow Huang into a new entity, or remain with Nvidia under a financing structure without him.
Would $50 Billion Be Enough?
If Huang were to launch a new company with $50 billion in initial backing, that figure would need to be assessed against the current funding environment for AI chip startups.
For comparison, British AI inference chip company Fractile completed a $720 million Series C funding round in November at a $7 billion valuation. South Korea’s FuriosaAI reached a valuation of $2.9 billion following its latest funding round. These are among the more closely watched companies in today’s AI chip startup sector, where funding rounds range from hundreds of millions to billions of dollars. A $50 billion starting fund would be many times larger.
The scale suggests that Goswami’s prediction points to an entity capable of reshaping competition from the outset, rather than a typical AI chip startup. A $50 billion funding base could support investment across the entire chain—from chip design and software development to securing foundry capacity and building a customer ecosystem—without repeated early fundraising rounds that dilute control.
However, capital alone does not create a competitive barrier. Fractile’s roadmap, which involves stacking DRAM and logic chips in 3D to deliver greater bandwidth and speed than conventional HBM, points to competition between different technological approaches. Success is about more than spending more money on the same approach. If Huang were to establish a new company, he would enter a market where Nvidia holds more than an 80% share, while customers and investors are actively seeking alternatives. Microsoft has publicly expressed a desire for more AI chip options, and new entrants such as Fractile and FuriosaAI continue to attract funding.
That is precisely why Goswami’s reference to “friends who have supported him all along” carries weight. What Huang would need to recreate in a new entity is the network he has built over three decades: investors willing to back him, customers willing to try new chips, and supply chain partners willing to allocate production capacity to a new entrant.