Until recently, the prevailing narrative around Nvidia went something like this: In the early years of the AI boom, Nvidia was the sole provider of cutting-edge GPUs, a position that proved incredibly lucrative as the industry expanded. Over the last few years, however, hyperscalers such as Amazon and Google have begun developing their own chips, meaning Nvidia is no longer the only option. That shift has prompted many investors to question just how lasting its competitive edge really is.

It’s a compelling story—and largely accurate. After seeing its market cap grow tenfold between early 2023 and mid-2025, Nvidia’s stock has followed a more subdued path over the past year, fueled by worries about GPU competition. But since the company’s earnings report on Wednesday, a fresh narrative has emerged, and investors are beginning to recognize that Nvidia’s strengths extend well beyond the GPU itself.

As AI compute scales up to gigawatt levels, orchestration has become an increasingly intricate challenge. Unsurprisingly, Nvidia has built much of the state-of-the-art hardware needed to tackle it, giving the company a significant edge in the systems surrounding the GPU—even as it faces fiercer competition on the GPUs themselves. For all the talk about compute becoming a commodity, operating a megascale data center at peak efficiency remains extraordinarily difficult. And as deployments grow larger and faster, that difficulty only intensifies.

Rack by Rack

You can see some of this just by examining what Nvidia is actually selling. The company is currently rolling out its Vera Rubin architecture, which pairs the Rubin GPU with a suite of other components, including the Vera CPU, the Groq 3 LPX inference accelerator, and corresponding racks for storage and networking. Over the past week, I’ve been speaking with people at Nvidia about what these systems actually do, and the findings have been eye-opening.

Like the Rubin GPU itself, these are highly specialized systems, but instead of processing tokens, they ensure everything outside the GPU operates as efficiently as possible. If the GPU is the engine, these are the rest of the car. The Vera CPU, in particular, is designed to tackle the problem of data orchestration.

“Vera is important because there’s only so much memory that you can put in a single server or any sort of compute platform,” Jason Hardy, Nvidia’s VP of storage technology, told me. As data centers have ramped up computing power, memory capacity has also scaled—which is why companies like Micron have thrived in the second wave of the infrastructure boom. But getting that data to the GPU at the right moment isn’t straightforward. As companies push to drive tokens-per-watt ever lower, they’re realizing how critical that kind of traffic direction truly is.

“We saw upwards of 3x improvement in these operations, where the Vera CPU is allowing for acceleration,” Hardy said. “So now we can use our flash to its fullest potential, because we can get all that performance out of it without bottlenecking.”

You can see versions of the same problem outside of Nvidia. When OpenAI developed its Jalapeño chip, a major focus was sidestepping these challenges entirely by minimizing the amount of data that needs to be moved around. “We designed Jalapeño to minimize data movement and communication delays,” the company said in a blog post earlier this month. “Its large domain allows the entire workload to remain within one connected system, minimizing data movement and helping the complete request stay fast and efficient from beginning to end.”

It’s a different approach—avoiding data movement altogether by handling a workload within a single integrated chip. But the underlying logic is the same: boosting efficiency through smarter traffic control rather than simply adding more processor cycles. That, in turn, opens up an entirely new layer of infrastructure for companies to compete over.

This new emphasis on data orchestration isn’t automatically a win for Nvidia. The company will have to compete with rival chipmakers and hyperscalers just as it has with GPUs. But the competition has shifted to a new level, where building a rival GPU matters less than being able to make the entire system work efficiently. And at least in the early stages, Nvidia appears to hold a commanding lead.