Nvidia’s CEO Jensen Huang announced a new chip planned for 2026 three months after launching a new model.
Nvidia’s CEO Jensen Huang announced a new chip model to be released in 2026 at the COMPUTEX tech conference in Taipei, Taiwan. The announcement came three months after the latest model presentation: a new chip to be released by 2024 dubbed “Blackwell”.
The current Hopper chips are Nvidia’s flagship products, with the likes of Meta, Google, and OpenAI as clients. Although prices are not publicly available, one Hopper chip is rumored to retail at around $30,000.
The company is at the forefront of the ongoing artificial intelligence (AI) boom, manufacturing chips that power popular software like OpenAI’s ChatGPT and Google’s Gemini.
This allowed Nvidia to become the uncontested global leader in chip design. Since ChatGPT’s release in late 2022, Nvidia’s stock price surged by almost 3,000% to $1,133. This made Nvidia the world’s third-most valued company at $2.79 trillion, falling behind only Microsoft and Apple.
Nvidia posted a slew of better-than-expected quarterly earnings since the beginning of the AI frenzy. In the latest quarterly report on May 22, Nvidia posted a 427% revenue jump since last year at $26 billion. Net profits for the latest quarter amounted to $14.9 billion, beating market estimates.
Looking ahead
Despite Nvidia’s overly dominant market position, some of its rivals are poised to challenge it. Specifically, AMD and Intel, formerly the world’s largest chipmaker, are looking at ways to enter the AI chip market.
Intel recently received a $20 billion government grant to build chip manufacturing plants on US territory. Nvidia outsources chip manufacturing to Taiwan Semiconductor Manufacturing Corp. (TSMC), placing its operations in the delicate balance of US-China geopolitical tensions.
In order to outrun the competition, Nvidia plans to expand operations and offer a wider choice of powerful AI chips.
The new “Blackwell” model will be twice as powerful as Hopper. The other model planned for 2026 announced on Sunday, dubbed “Rubin”, is expected to be even faster and more powerful.
In general, Huang announced that Nvidia will increase chip releases to “once a year” instead of the twice-a-year strategy implemented so far. “Our basic philosophy is very simple,” Huang said during the announcement, “build the entire data-center scale, disaggregate and sell to you parts on a one-year rhythm, and push everything to technology limits.”
“Today, we’re at the cusp of a major shift in computing,” Huang added. “With our innovations in AI and accelerated computing, we’re pushing the boundaries of what’s possible and driving the next wave of technological advancement.”
