By Aditya Soni and Anhata Rooprai
June 8 (Reuters) – Nvidia’s entry into the AI PC market with its RTX Spark superchip last week is less a breakthrough for regular users than a high-stakes bet that a largely unproven concept can find wider appeal, analysts said.
At the Computex trade show in Taiwan, the chipmaker pitched a future where laptops run large AI models locally and act as personal digital agents, no cloud needed.
It’s a claim that PC makers HP and Dell have made for nearly three years now, only to be met with skepticism from Wall Street and consumers, with high prices outweighing tangible benefits.
Nvidia, though, appears to be selling a different version of the AI PC than what exists today, one aimed more at developers and content creators who have long favored Apple’s high-end MacBook Pros. Six companies – Microsoft, Asus, HP, Lenovo, Dell, and MSI – will make PCs using the chip. These stocks surged after Nvidia’s announcement on June 1.
“RTX Spark doesn’t make traditional PCs obsolete. It creates a new category between the workstation and the AI server,” said Kevin Hein, analyst at Tirias Research.
The chip combines a central processor, graphics engine and up to 128 gigabytes of unified memory, making it capable of running large AI models locally, something current AI PCs cannot do at scale. Nvidia says it could reshape how people interact with computers, with AI agents handling tasks such as generating videos or debugging code.
Existing AI PCs, marketed heavily over the past two years, have centered on modest features like transcription or image editing and have failed to drive meaningful sales for the device manufacturers and their partners such as Arm and Qualcomm.
COST BARRIERS LOOM
A premium price and a memory chip crunch, which has already driven up device cost, are likely to limit RTX Spark devices to niche adoption, analysts said.
The cost “won’t deter all the big computer makers from working with Nvidia on this, but the bulk of PC sales for the next several years will still be more traditional Windows-based PCs with chips from Intel, AMD and Qualcomm,” said Bob O’Donnell, president at TECHnalysis Research.
HP and Dell stock had been climbing even before the launch of Nvidia’s superchip, up 18% and 223%, respectively, so far this year. But that rally has been driven less by AI PCs and more by a wave of corporate upgrades to Windows 11, as well as booming demand for AI infrastructure, particularly for Dell.
In its latest quarter, HP warned of a sharp decline in the PC market in the latter half of the year. The company did note strong demand for AI PCs, especially from enterprise customers, although the overall PC business recorded shrinking sales.
The outlook for PC sales looks dour this year, with IDC estimating global PC shipments to decline 11.3% in 2026.
COMPETING WITH APPLE
It is unclear if devices powered by the Nvidia chip will outperform Macs. Nvidia said details on battery life and other metrics would be shared closer to the launch of the products this fall.
Still, the Nvidia-powered laptops could make Windows machines competitive with Macs for the first time on memory bandwidth, a key bottleneck for AI software that constantly shuttles data back and forth between a machine’s processor and memory, adding latency.
That brings it closer to Apple’s in-house chips, which have bundled unified memory since 2020.
“I expect some companies will take the leap to test out the long-term viability of on-device inferencing,” said Tom Mainelli, a group vice president at IDC.
(Reporting by Aditya Soni amd Anhata Rooprai in Bengaluru; Editing by Sayantani Ghosh and Anil D’Silva)
Comments