Nvidia and AMD reveal upcoming AI processors amid intense rivalry.
Nvidia CEO Jensen Huang announced on Sunday that the company plans to launch their most advanced AI chip platform, called Rubin, in 2026. Rubin will replace the Blackwell platform, which was announced in March and hailed as the "world's most powerful chip."
The Rubin will include new graphics processing units (GPUs), a new central processing unit (CPU) called Vera, and advanced networking chips, according to Huang. He made this statement during a speech at National Taiwan University in Taipei.
"We're on the brink of a significant change in computing," Huang told the audience ahead of the 2022 Computex, an annual technology trade show in Taiwan. "The intersection of AI and accelerated computing is poised to shape the future."
Huang shared a plan for new semiconductors released at a one-year interval. Investors have been driving up shares of chip companies involved in the generative AI boom. Nvidia's stock has more than doubled in the past year.
Analyst Richard Windsor of Radio Free Mobile, a research company focusing on the digital and mobile ecosystem, said, "Nvidia clearly wants to keep its dominance for as long as possible, and in the current generation, there's nothing that can challenge that." Nvidia accounts for approximately 70% of AI semiconductor sales.
However, competition is heating up, with major rivals AMD and Intel introducing new products to challenge Nvidia's market leadership.
AMD CEO Lisa Su revealed her company's newest AI processors in Taipei and a plan to create new products over the next two years. She announced the MI325X accelerator would be available in the fourth quarter of 2022.
"AI is our top priority, and we're at the start of an incredibly exciting time for the industry," Su said. "Our new product family will be released on an annual cadence, meaning a new product every year."
The upcoming chip will succeed the MI300 and feature more memory, faster memory bandwidth, and improved computer performance, Su added. AMD will release a new product family each year, with the MI350 slated for 2025 and MI400 the following year.
Intel is also expected to release its next-generation desktop CPUs, Arrow Lake, in the fourth quarter of this year. Its CEO, Patrick Gelsinger, will speak at Computex on Tuesday.
Both Nvidia and AMD, which are led by Taiwanese CEOs, were previously known for selling graphics processors for gaming. Although they still compete in that market, their GPUs are now used to power AI technology like ChatGPT.
"AI is our top priority, and we're at the beginning of an incredibly exciting time for the industry," Su concluded. "We unveiled the MI300X last year, offering industry-leading inference performance, memory size, and computing capabilities. Now, we're expanding our roadmap because we'll be releasing a new product family annually, starting with the MI350 in 2025 and the MI400 in 2026."
Read also:
The business strategy of Nvidia and AMD heavily involves the tech sector, as both companies are actively developing and launching advanced AI processors to maintain their market dominance. Nvidia's upcoming AI chip platform, Rubin, will include advanced networking chips, a new GPU, and CPU, Vera, further cementing their position in the tech-focused AI semiconductor market.
