Nvidia unveils AI-powered PC chips to challenge Intel, AMD

Nvidia, which has become the world’s most valuable company through its dominance in the AI chip market for data centers, is now expanding into the personal computer processor sector — a market long dominated by Intel, AMD, Qualcomm and Apple.

During a keynote speech at the Computex technology conference in Taiwan on Monday, Nvidia CEO Jensen Huang introduced the company’s new N1X processor, developed in collaboration with Microsoft, News.Az reports, citing CNBC.

The chip will power Nvidia’s new RTX Spark superchip platform, which is set to launch this fall in a new generation of Windows PCs produced by Microsoft, Dell, HP, ASUS, Lenovo and MSI.

“Microsoft and Nvidia are going to reinvent the PC,” Huang said during the presentation. “This is the first completely re-engineered, reinvented line of PCs that has happened in 40 years.”

According to an Nvidia spokesperson, the company’s initial rollout plan includes more than 30 laptop models and 10 desktop systems using the new processor.

Huang compared the significance of the development to the transformation of mobile phones into smartphones, saying the new devices are designed to support advanced agentic AI capabilities across all systems.

The debut processor combines two of Nvidia’s flagship chip technologies along with 128 gigabytes of unified memory. The RTX Spark platform integrates one of Nvidia’s Blackwell graphics processing units (GPUs) with the company’s new Arm-based N1X central processing unit (CPU), which was custom-designed in partnership with Taiwanese semiconductor firm MediaTek.

The launch could significantly reshape the PC industry, which is already undergoing major changes driven by the rapid growth of artificial intelligence technologies. Arm-based processors, including Nvidia’s new chip, are increasingly challenging the traditional x86 processors developed by Intel and AMD, while the global CPU market is expanding rapidly into what Huang described as a potential $200 billion industry.

Nvidia previously told CNBC in February that CPUs were becoming a major bottleneck as demand for agentic AI workflows increased. In March, the company introduced a full server rack powered by its Vera CPUs for data centers.

While GPUs are essential for training large AI models through parallel computing, CPUs play a critical role in managing and distributing data across AI systems and agents.

Nvidia confirmed that the new PC processor will be manufactured using Taiwan Semiconductor Manufacturing Company’s advanced 3-nanometer process technology, which is currently available only in Taiwan.

Speculation about Nvidia entering the Arm-based PC chip market has circulated for several years. Reuters reported in 2023 that Nvidia had been developing a PC processor as part of Microsoft’s broader strategy to encourage companies to produce Arm-based chips for Windows devices.

An Nvidia spokesperson said the company has worked with Microsoft on the project for “many, many years,” describing the new processor as significantly more capable, efficient and higher-performing than traditional x86-based alternatives.

Intel pioneered the x86 instruction set architecture in the 1970s and on Monday also unveiled its latest Xeon 6+ data center CPUs at Computex.

In recent years, many technology companies have shifted toward Arm’s energy-efficient architecture, which gained mainstream attention with the launch of the original iPhone in 2007.

Apple now uses Arm-based chips in its Mac computers and introduced its latest M5-powered MacBooks earlier this year. Arm also launched its first in-house CPU in March, while AMD is reportedly developing its own Arm-based PC processor.

The first laptops powered by Nvidia’s RTX Spark platform will reportedly measure as thin as 14 millimeters and target premium consumers. The devices will also be available in compact desktop versions.

Although Nvidia plans to expand the platform to additional price ranges in the future, the company said the initial focus is on creators, AI developers and gamers seeking lightweight laptops and compact desktop systems.

Nvidia said more detailed performance benchmarks will be released closer to the official launch later this year. For now, the company stated that RTX Spark delivers graphics performance roughly comparable to Nvidia’s RTX 5070 laptop GPU.

At Computex, Huang also announced that Nvidia’s Vera CPU platform for data centers has entered full-scale production. He said the company is manufacturing millions of the processors for what he described as an entirely new market segment.

The Vera CPUs are expected to become available this fall, with early customers including Anthropic, OpenAI, xAI, Dell, Oracle and CoreWeave.

“This is going to be our new major growth driver,” Huang said. “These CPUs are going to be both performant, but they also have to be extremely energy efficient.”

Ian Buck, Nvidia’s vice president of hyperscale and high-performance computing, said faster CPUs have become critical for maintaining AI infrastructure and improving the efficiency of AI systems.

According to Buck, Vera CPUs can generate AI tokens 1.8 times faster than current x86 processors, helping create smarter AI agents and increasing overall data center productivity.

News.Az 

By Nijat Babayev