Annabel Chee/Bloomberg/Getty Images
Nvidia co-founder and CEO Jensen Huang speaks in Taipei on June 2, 2024.
Taipei/Hong Kong
CNN
—
Nvidia and AMD have each launched their next-generation artificial intelligence (AI) chips in Taiwan as the three-way race with Intel intensifies.
Nvidia (NVDA) CEO Jensen Huang said on Sunday that the company will launch its most advanced AI chip platform, called Rubin, in 2026.
The Rubin platform, a successor to Blackwell's data center chip supplier, was only announced in March, when Nvidia called it “the most powerful chip in the world.”
Speaking at National Taiwan University in Taipei, Huang said Rubin will be equipped with a new graphics processing unit (GPU), a new central processing unit (CPU) called “Vera” and advanced networking chips.
“We are currently at a major turning point in computing,” Huang told an audience ahead of the opening of COMPUTEX, Taiwan's annual tech trade show. “The convergence of AI and accelerated computing will redefine the future.”
He revealed a roadmap for new semiconductors that will be released on an “annual cycle.”
Investors are riding the generative AI boom to boost semiconductor company shares, with market leader Nvidia's shares more than doubling over the past year.
“Nvidia is clearly looking to maintain its dominance for as long as possible and I don't see anything in this generation that can compete with it,” said Richard Windsor, founder of Radio Free Mobile, a research firm focused on the digital and mobile ecosystem.
Nvidia accounts for roughly 70% of AI semiconductor sales, but competition is heating up, with major rivals AMD (AMD) and Intel (INTC) introducing new products to challenge Nvidia's dominance.
On Monday, AMD CEO Lisa Su was in Taipei to unveil the company's latest AI processors as well as its new product development plans over the next two years.
The company's next-generation MI325X accelerator is expected to be available in the fourth quarter of this year, she said.
The global race to create generative AI applications is driving a surge in demand for cutting-edge chips used in data centers to support these programs.
Nvidia and AMD, both run by the same Taiwanese family CEO, were once best known to gamers for selling GPUs that display video game images and bring the games to life.
While the two companies still compete in that space, their GPUs are now also being used for generative AI, the technology behind newly popular systems such as ChatGPT.
“AI is a top priority for us and we are at the beginning of a very exciting time for the industry,” Su added.
“We launched the MI300X last year with leadership inference performance, memory size and compute power, and now we're expanding our roadmap to an annual cadence, releasing a new product family every year,” she said.
Xu further stated that the new chip will be the successor to the MI300, featuring larger memory capacity, faster memory bandwidth and improved computing performance. The company will launch a new product series every year, with the MI350 expected to launch in 2025 and the MI400 a year later.
Intel is also set to launch its next-generation desktop CPUs, Arrow Lake, in the fourth quarter of this year. The company's CEO, Patrick Gelsinger, is scheduled to speak at Computex on Tuesday.
