Nvidia, a company previously recognized for its indispensable graphics processing units (GPUs) in the artificial intelligence (AI) sector, has expanded its product line by introducing its own central processing unit (CPU) called Vera. This move positions the chip giant to challenge established industry leaders, Advanced Micro Devices (AMD) and Intel, in the rapidly growing AI server market. The company recently released detailed specifications and technical benchmarks for Vera, providing prospective customers with the information needed to fully evaluate the chip.
Strategic Shift into the CPU Market
While Nvidia established its market dominance through its GPUs—the primary chips used for creating and deploying AI—it is now shipping its own CPUs. These processors give cloud service providers an alternative to chips manufactured by AMD and Intel, creating a new competitive arena within AI servers. The company provided new details on its Vera data center CPU, stating that samples had been delivered to major clients, including SpaceX, Anthropic, and OpenAI, in June.
Nvidia’s investment in the CPU space exemplifies a strategy aimed at vertically integrating its systems. Rather than merely selling individual chips, the company intends to sell complete, fully integrated computing racks. According to Nvidia, this strategy will enable engineers to extract greater performance from their GPUs, helping the company’s systems remain the preferred choice for leading AI laboratories as competition from AMD and custom chips intensifies.
Historically, CPUs were the most critical and expensive component in a server before the AI boom. However, the rise of “agentic AI”—which can function autonomously with minimal human intervention—has increased the importance of the CPU, as it is necessary for managing and supplying data to AI agents. This shift has been noted by financial markets, with CPU incumbents AMD and Intel showing strong performance in 2026.
Technical Specifications and Performance Advantages
Nvidia stated that Vera is unique because it is the first server CPU the company has engineered from the ground up, rather than utilizing an existing design from Arm. The chip was specifically designed to resolve bottlenecks encountered by AI agents. Nvidia claims that Vera offers 50% better performance for AI agents compared to x86 chips, the type used by Intel and AMD. The company explained that while previous CPU designs from Intel and AMD focused on core count, Vera and the associated Olympus core emphasize single-core speed instead.
Hannah Coutand, a Vera product marketer at Nvidia, noted during a presentation that the chip prioritizes high memory bandwidth, per-core speed, and low latency. She explained this focus allows AI agents to quickly return control to the GPUs, ensuring that these “very expensive and a highly valuable asset in the AI factory” remain highly utilized.
The CPU is designed to consume between 250 watts and 450 watts of power and supports a substantial amount of low-power memory, up to 1.5 terabytes per chip. Nvidia plans to sell the chip both standalone and paired with its GPUs, including as a liquid-cooled rack of 256 Vera chips, or in a configuration with two Vera chips within a single server. This combination is branded as the Vera Rubin system.
Market Outlook and Industry Analysis
Analysts have taken notice of Nvidia’s effort, suggesting the company has created a new class of CPU that currently lacks a direct competitor from Intel or AMD. Karl Freund, founder of Cambrian AI Research, stated that Nvidia’s new CPU is intended for intensive AI tasks, not for traditional server functions like serving websites. However, Freund cautioned that convincing cloud providers to purchase the chips remains a challenge for the company.
Industry projections for the server CPU market are significant. Nvidia estimates the entire server CPU market could eventually reach $200 billion, contrasting with an earlier Bernstein estimate that placed the mature server CPU market at approximately $37 billion in 2025. Wolfe Research projected that the average selling price for a Vera chip would be around $5,000, forecasting that Nvidia would ship about 1.3 million units this year. Nvidia declined to provide any comments regarding its pricing structure.
Market analysts have also provided insights into the competitive landscape. Kevin Knox, an analyst from Gartner, suggested that AMD is currently the leading contender in enterprise AI server CPUs. Knox noted that while Intel reportedly holds a substantial market share, AMD is actively increasing its market penetration and maintains strong connections with hyperscale cloud providers, crediting AMD’s success in building its chip ecosystem.
