Nvidia Unveils RTX Spark Chip and Vera CPU, Powers HPE’s New AI Server
NVDA•Nvidia unveiled the RTX Spark PC chip at Computex, enabling local AI agent workloads in laptops and desktops, with shipments starting this fall. Simultaneously, Nvidia introduced its Vera CPU—offering 2x efficiency over x86—integrated into HPE’s new DL394 Gen12 servers, set for NYSE deployment in fall 2026.
1. PC Chip Launch
On June 1 at Computex in Taiwan, Nvidia unveiled the RTX Spark PC chip, designed to run AI agents locally on laptops and desktops without relying on cloud infrastructure. The chip, developed in collaboration with Microsoft over three years, aims to transform traditional app-centric PCs into agentic AI personal computers and will begin shipping this fall.
2. Vera CPU Unveiled
Nvidia also introduced its Vera central processing unit, built specifically for agentic AI workloads and claiming twice the efficiency and faster task completion compared to x86 architectures. Early adopters of Vera CPUs include OpenAI, Anthropic and SpaceX, positioning Nvidia to enter a potential $200 billion market for AI-tailored processors.
3. HPE Server Integration
Hewlett Packard Enterprise launched the ProLiant Compute DL394 Gen12 server featuring Nvidia’s Vera CPUs, offering industry-leading memory bandwidth and low-latency performance for AI and data processing workloads. The New York Stock Exchange plans to deploy these servers in fall 2026 to scale capacity above 1.1 trillion daily messages and optimize latency for resilient, AI-ready market infrastructure.
