Nvidia RTX Spark Launch Triggers AI PC Chip War, Slashing Qualcomm Shares by 8.8%

Nvidia RTX Spark Launch Triggers AI PC Chip War, Slashing Qualcomm Shares by 8.8%

The battle for artificial intelligence silicon supremacy has expanded into a multi-front war across both consumer client devices and data center infrastructure. At Computex 2026 in Taipei, NVIDIA launched a direct assault on the premium PC market with its TSMC 3nm RTX Spark™ Superchip, developed in collaboration with Microsoft to power high-end Windows on Arm laptops. This announcement triggered a sharp 8.8% decline in Qualcomm (QCOM) shares on June 2, 2026, closing at $228.99, as investors feared direct competition to Qualcomm's Snapdragon X Elite series.

Nvidia's Data Center Expansion: The Vera CPU and SK Hynix Alliance

NVIDIA is not just squeezing Qualcomm in the PC market; it is also launching a direct assault on the traditional data center CPU market. At Computex, NVIDIA announced that its next-generation Vera Rubin server platform is in full production, with shipments scheduled for Fall 2026.

The cornerstone of this platform is the Vera CPU, an 88-core standalone microprocessor based on NVIDIA's Olympus CPU core and built on TSMC's 3nm process. On June 7, 2026, NVIDIA CEO Jensen Huang announced in Seoul that the Vera CPU will utilize SK Hynix DRAM chips:

"We introduced Vera CPU, which is a revolutionary CPU, and it will also use SK Hynix’s DRAM."

Vera represents NVIDIA's first standalone data center microprocessor, designed to compete head-to-head with Intel's Xeon, AMD's Epyc, and cloud hyperscalers' custom silicon (such as Amazon's Graviton). Furthermore, NVIDIA is expanding its reach into client-side professional workstation developers with the DGX Station for Windows, powered by the GB300 chip (72 Grace cores, Blackwell Ultra GPU, and 252GB HBM3e) requiring 1600W of power, slated to launch in Q4 2026.

Qualcomm's Counter-Offensive: The "Dragonfly" Brand

In a direct bid to protect its turf and enter the lucrative AI infrastructure space, Qualcomm CEO Cristiano Amon declared 2026 the "Year of AI Agents" and officially launched "Dragonfly", Qualcomm's new brand for data center AI inference products.

• The Strategy: Dragonfly signifies Qualcomm's entry into the data center market, offering CPUs, AI inference accelerators (including the upcoming AI200 and AI250 chips), and custom silicon.
• What to Watch: Qualcomm is scheduled to hold a major Investor Day in New York City on June 24, 2026, where it will detail the technical specifications, hyperscaler partnerships, and physical AI/robotics roadmap for the Dragonfly suite.

This sets up an intense competitive collision: Qualcomm is attempting to break into Nvidia's data center stronghold via Dragonfly, just as Nvidia's RTX Spark and Vera CPU threaten Qualcomm's client PC and server ambitions.¹