Nvidia Will Now Make New AI Chips Every Year

Nvidia, fresh off a staggering $14 billion profit in a single quarter thanks to its AI chips, is revving up its innovation engine. The company will now design and release new AI chips annually, instead of its previous biennial schedule, according to CEO Jensen Huang.

“I can announce that after Blackwell, there’s another chip. We’re on a one-year rhythm,” Huang revealed during Nvidia’s Q1 2025 earnings call.

A New Cadence for Innovation

Traditionally, Nvidia has introduced a new architecture every two years. This pattern brought us Ampere in 2020, Hopper in 2022, and Blackwell in 2024. The industry-favored H100 AI chip came from the Hopper architecture, and Blackwell has given rise to the B200. These architectures also find applications in Nvidia’s gaming and creator GPUs.

However, earlier this month, analyst Ming-Chi Kuo suggested that the next architecture, dubbed "Rubin," is slated for 2025, potentially introducing the R100 AI GPU as soon as next year. Huang's recent comments support this forecast, indicating that Nvidia is indeed accelerating its release cycle.

Accelerating Across the Board

Huang emphasized that this rapid pace will apply to all Nvidia’s product lines. “We’re going to take them all forward at a very fast clip,” he stated. This includes not just GPUs, but also CPUs, networking NICs, and switches. “New CPUs, new GPUs, new networking NICs, new switches... a mountain of chips are coming,” he added.

Seamless Transitions

One of the key points Huang highlighted was the backward compatibility of Nvidia’s new AI GPUs. When asked about the transition from Hopper GPUs to Blackwell GPUs, Huang explained that the new generations are electrically and mechanically backward-compatible and run the same software. This allows customers to “easily transition from H100 to H200 to B100” in their existing data centers.

Driving Demand

Huang also addressed the immense demand for Nvidia’s AI GPUs. “We expect demand to outstrip supply for some time, as we transition to H200, as we transition to Blackwell,” he said. This demand surge is driven by companies eager to save money and enhance profitability by getting their AI infrastructure online as quickly as possible.

The FOMO Factor

Huang also made a compelling argument for the urgency in adopting Nvidia’s cutting-edge technology. He highlighted the competitive advantage of being the first to announce groundbreaking AI capabilities versus being second with marginally better performance. “The next company who reaches the next major plateau gets to announce a groundbreaking AI, and the second one after that gets to announce something that’s 0.3 percent better,” he quipped. “Do you want to be the company delivering groundbreaking AI, or the company, you know, delivering 0.3 percent better?”

Expanding Horizons

Interestingly, Nvidia’s CFO noted that the automotive sector is set to become the company’s “largest enterprise vertical within data center this year.” Tesla, for instance, has purchased 35,000 H100 GPUs to train its full-self driving system. Meanwhile, "consumer internet companies" like Meta continue to be strong growth drivers. Meta plans to have over 350,000 H100 GPUs in operation by the end of the year, underscoring the massive scale of demand.

Conclusion

Nvidia’s decision to switch to a yearly release cycle for its AI chips marks a significant shift in the tech industry’s landscape. By accelerating innovation across all its product lines, Nvidia is positioning itself to maintain and expand its leadership in AI technology. This strategic move not only meets the growing demand for AI infrastructure but also sets a new standard for innovation cadence in the tech industry.

Stay tuned with Kushal Bharat Tech News for more updates on Nvidia’s groundbreaking advancements and other exciting tech news.