Nvidia, AMD, and Intel Are Fighting the Most Important Tech War in Decades

On a weekday afternoon in Santa Clara, there is a certain silence that is only broken by the hum of HVAC systems cooling buildings occupied by engineers who sincerely think they are creating the future. Employees clutching laptops and badges can be seen walking around the AMD campus with the calm assurance of those whose stock options have recently performed exceptionally well. That scene would have been unimaginable ten years ago.

In 2014, Lisa Su took over AMD, which was valued at less than $3 billion and regarded as a charitable survivor. It is currently above $330 billion and continues to rise. At the time, the shift from gaming graphics cards to data-center silicon was not immediately apparent. As is usually the case, it appears clear now.

Naturally, Nvidia is the business that everyone is attempting to acquire. Wearing his now-uniform leather jacket, Jensen Huang has been gathering clients for the past two years in a manner similar to that of some vintage watch collectors. However, an intriguing development recently occurred. Earlier this year, Huang began discussing CPUs, the very chips that Intel and AMD have built their identities around, on an analyst call. His casual ambition caused seasoned chip watchers to pause. He even suggested that Nvidia might grow to be “one of the largest CPU makers in the world.” It’s the kind of comment you would brush off if it came from someone else.

In the meantime, Intel is attempting to make a comeback, which is the hardest thing in technology. After receiving antitrust approval for their investment in SambaNova, Lip-Bu Tan’s team recruited Qualcomm’s Alex Katouzian to lead a new Client Computing and Physical AI division. Quarter over quarter, the foundry margins increased by 150 basis points. Small numbers, but the kind of small numbers that are more significant than the large ones in a turnaround story. Although belief and proof are two different things, investors appear to think Intel may finally be stabilizing.

Three companies aren’t really at the center of the deeper story. It’s about the evolution of computing itself. For many years, everything was an accessory and the CPU was the brain. Then, because video game math turned out to look a lot like AI math, the GPU almost unintentionally rewrote that hierarchy. The balance is now shifting once more as AI agents write code and parse documents on their own. According to Ben Bajarin of Creative Strategies, the CPU to GPU ratio in AI servers may tilt back or collapse toward parity. As of yet, no one knows.

Observing this develop, it’s odd how intimate it has become. Lip-Bu Tan, Jensen Huang, and Lisa Su are all members of the Taiwanese-American engineering diaspora. In certain situations, they have known one another for years. In keynote addresses and SEC filings, the rivalry has the tension of a family dispute. It’s difficult to ignore the fact that the individuals choosing the architecture of contemporary intelligence all grew up in the same tough environment of Bay Area chip design in the 1990s.

It’s unclear whether Nvidia’s CPU aspirations are genuine, whether AMD achieves its purported trillion-dollar valuation, or whether Intel’s turnaround is sustained. The war has transcended the hype, that much is certain. Now, it comes down to execution, profit margins, and which business can deliver on its commitments. To put it another way, the difficult part is just getting started.