A $5 million check in 1995, a $4 trillion market cap in 2026: Huang revisits the Sega deal that kept Nvidia alive

Nvidia chief executive Jensen Huang met former Sega chief executive Shoichiro Irimajiri and veteran game designer Yu Suzuki in Tokyo on 15 July 2026 to thank them, in person, for a $5 million investment placed with Nvidia in 1995, an outlay the chipmaker has said kept the company solvent through its first year in the graphics-accelerator business.

The reunion, first flagged by X account @unusual_whales, is part gratitude and part signal. Three decades on, Nvidia's market value stands at roughly $4 trillion, and the same Japanese executives who once bet on a near-failure are now seated across from the architect of the silicon stack that trains the world's largest artificial-intelligence systems. The optics matter: Tokyo is one of the first stops on Huang's current Asia circuit, and Japan's contract-manufacturing base, led by Tokyo-listed partners including Disco, Advantest and Tokyo Electron, has become central to Nvidia's near-term supply plan.

The trip is a reminder that the company's present dominance has a longer memory than the press cycle usually allows.

What the 1995 deal actually was

By the close of 1994, Nvidia had shipped its first product, the NV1, a combined 2D-and-3D graphics accelerator, into a PC market that was rapidly standardising around Microsoft Direct3D. The chip was, by Nvidia's own later telling, the wrong bet. Sega had agreed to use Nvidia silicon in a forthcoming console project, but in mid-1995 Sega switched to a rival design for what would become the Saturn. Irimajiri, then Sega's chief executive, and Yu Suzuki, the designer behind Sega arcade franchises including Virtua Fighter and Shenmue, nonetheless directed a roughly $5 million investment into Nvidia, a move Huang has credited over the years with keeping payroll alive long enough for the company to ship its next-generation RIVA 128 in 1997.

The sum, $5 million in 1995 dollars, is the kind of figure that gets rounded up in anniversary copy. It is also the figure Nvidia itself has used repeatedly when the company tells its origin story in talks and on its investor-relations site.

Why this matters now

The Tokyo stop lands at a moment when Japan's role in the AI supply chain is no longer decorative. Tokyo Electron supplies critical deposition and etch tools, Advantest makes the testers that screen every advanced GPU that comes off the line, and Disco Corp's precision wafer-dicing equipment has become a quiet choke point for advanced packaging. Government-backed JASM, the foundry joint venture in Kumamoto led by Taiwan Semiconductor Manufacturing Company (TSMC), is in early production and is being marketed partly to Nvidia-adjacent customers. The economics of all of that are partly Nvidia's economics; any constraint at the back end is a constraint on training-fleet buildouts.

Seen from Tokyo, a courtesy call on Irimajiri and Suzuki is also a courtesy call on an industrial base that is in the middle of a once-in-a-generation capacity buildout, and on a Japanese investment community that is being courted to participate in it.

The counter-read

The softer framing of the trip is, of course, simply a thank-you. The harder framing is that Huang is using a known origin story to do commercial work in real time: signalling to Japanese conglomerates and to the Tokyo stock exchange that Nvidia intends to keep buying, hiring and partnering inside Japan. A third framing, less flattering, treats the anecdote as a piece of founder mythology, polished for investors: the near-death pivot, the loyal customers, the lucky break. All three readings are consistent with the same set of facts. The one to watch is whether the Tokyo leg produces a named Japanese partner announcement before Huang leaves the region. That is the difference between a nostalgia tour and a sourcing visit.

The stakes behind the optics

For Japan, the upside is a larger share of the value-add inside AI hardware, a share that has so far gone mostly to Taiwan, South Korea and the United States. Tokyo's bet, expressed through METI subsidies and JASM equity, is that the next cycle of accelerator production gets localised. For Nvidia, the upside is redundancy: another set of credible back-end suppliers at a moment when capacity, not demand, is the binding constraint on the AI buildout. For Irimajiri and Suzuki, the upside is, simply, that a 1995 decision is once again legible to a global audience, and that the company it kept alive remembers them by name.

The one thing the public record still does not specify is whether Sega's 1995 investment was structured as straight equity, a convertible instrument or a development advance that was later converted, a detail that would change its present-day accounting weight but not its present-day symbolic one.

Monexus read this against a single X thread (above) and corroborated the named figures against Nvidia's own publicly told origin story and against standard references on the 1995 Sega-Nvidia relationship. Where a figure could not be cross-checked, it was omitted rather than inferred.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://x.com/unusual_whales/status/1944712940416483509
• https://en.wikipedia.org/wiki/Sega
• https://en.wikipedia.org/wiki/Nvidia
• https://en.wikipedia.org/wiki/Yu_Suzuki