Nvidia CEO Jensen Huang speaks at Computex 2026 in Taipei, Taiwan, Monday, June 1, 2026. Computex is Asia’s largest electronics show, and in recent years has transformed from a PC show to a completely AI show.
Yifei Lin/Bloomberg via Getty Images
Nvidia has emerged as the world’s most valuable company by dominating the data center AI chip market. Now the company is expanding its reach into chips that serve as the main processors in personal computers, entering a field long dominated by Intel. advanced micro deviceQualcomm and Apple.
In his keynote speech at the Computex conference in Taiwan on Monday, Nvidia CEO Jensen Huang announced the new N1X processor, made in collaboration with Microsoft. It will be built into the new RTX Spark superchip and will be used by Microsoft, Dell and HP, ASUS, lenovo And MSI.
“Microsoft and Nvidia are reinventing the PC,” Huang said. “This is the first completely redesigned and reinvented PC series in 40 years.”
Nvidia’s initial plans include more than 30 laptops and 10 desktops equipped with the new chip, an Nvidia spokesperson said.
“This reinvention of the computer is as big an event as reinventing the phone into what we now know as the smartphone,” Huang said, pointing to the fact that agent AI will work on all new computers.
The debut PC processor is constructed from a fusion of two flagship types of Nvidia chips, plus 128 gigabytes of unified memory. It combines one of Nvidia’s Blackwell graphics processing units with a new Arm-based custom N1X central processing unit custom-designed by Taiwanese company MediaTek.
This reinvention of the computer is as big an event as the reinvention of the telephone into what we now know as the smartphone.
RTX Spark is a potential game-changer for the PC industry, which is already experiencing major changes due to the AI boom. While Arm-based processors like Nvidia are gaining ground over traditional x86 processors championed by Intel and AMD, the overall CPU market is exploding, becoming a $200 billion industry, according to Huang.
Nvidia told CNBC in February that CPUs are “becoming the bottleneck” as agent AI workflows proliferate. The following month, Nvidia announced an entire rack of Vera CPUs for data centers. Training large models requires massive amounts of parallel computation, which is great work for GPUs, but accessing that data and pushing it to multiple agents requires the more general computing that CPUs provide.
Nvidia’s new PC processors are Taiwan semiconductor manufacturing companyThe 3-nanometer technology is currently only available in Taiwan.
Expectations for Nvidia’s Arm-based PC chips have been building for years. Reuters reported that the company is working on developing a PC chip in 2023. microsoft The idea is to get companies to make Arm-based processors for computers. An Nvidia spokesperson said the company has been working with Microsoft on the chip for “years,” adding that it will be “much more powerful, capable, and efficient” than traditional x86 processors.
Intel was the first pioneer of the x86 instruction set, debuting in the 1970s. Intel announced its new Xeon 6+ data center CPUs at Computex in Taiwan on Monday.
More recently, many companies have been switching to Arm’s alternative power-efficient architecture, which first became mainstream with the original iPhone in 2007.
Apple currently makes Arm-based processors for its computers and will launch a more expensive MacBook series with the latest M5 chip in March. Arm also announced its first in-house CPU that same month, and AMD is also reportedly working on developing Arm-based PC chips.
The first laptops powered by Nvidia’s new chips will be a hefty 14 millimeters thick, and some smaller desktop models will also debut. RTX Spark will eventually expand to a variety of price points, but for now it’s targeted at creators, AI developers, and gamers who are “looking for an extremely thin and light laptop, slim laptop, portable laptop, or compact desktop,” Nvidia said.
Nvidia said it will release more performance metrics by the time the chip hits the market in the fall. For now, RTX Spark is “nearly equivalent” to Nvidia’s flagship RTX 5070 laptop GPU, according to a company spokesperson.
Huang also announced at Computex on Monday that Nvidia’s Vera CPUs for data centers are now running at full capacity. Huang said Nvidia is building millions of CPUs for “a market that didn’t exist before.” Vera will be available from fall. Early customers include Anthropic, OpenAI, SpaceX’s xAI, Dell, Oracle, and CoreWeave.
“This will be our new major growth driver,” Huang said. “These CPUs have good performance, but they also have to be very energy efficient so we can cram as many CPUs into the factory as possible without taking power away from token generation.”
“Fast CPUs have become essential to keeping AI factories running,” said Ian Buck, vice president of hyperscale and high-performance computing at Nvidia.
Buck said Vera can generate tokens 1.8x faster than current x86, “improving overall agent token performance, enabling agents to think smarter and longer, and ultimately generating more datacenter token revenue.”
WATCH: Exclusive First Look at Nvidia’s Vera Rubin
