Nvidia, a company that achieved significant market valuation due to the intense demand for its graphics processing unit (GPU)—the primary chip used for developing and deploying artificial intelligence—has expanded its offerings. The chip giant is now introducing its own central processing unit (CPU), named Vera, which aims to compete directly with established leaders AMD and Intel in the rapidly expanding AI server market.
Vera CPU and Strategic Market Expansion
Traditionally, Nvidia’s dominance was built around its GPUs. However, the company is now shipping its own CPUs, which cloud service providers may choose to utilize instead of products from Advanced Micro Devices (AMD) and Intel. On Tuesday, Nvidia provided new technical details concerning its data center CPU, Vera. This included specifications, architectural details, and benchmarks necessary for prospective customers to thoroughly evaluate the technology.
Nvidia representatives confirmed that Vera chips had already been delivered to major clients, including OpenAI, Anthropic, and SpaceX, in June. This expansion represents a strategic move into the CPU market, establishing a new competitive arena against the entrenched players, Intel and AMD. This effort also exemplifies Nvidia’s strategy to vertically integrate its systems, allowing the company to produce more components and technology within its own ecosystem annually.
Instead of merely selling individual chips, Nvidia aims to sell entire computing racks. The company stated that this strategy would assist engineers in maximizing the performance of their GPUs, helping to maintain Nvidia’s systems as the preferred choice for leading AI laboratories even as competition from AMD and specialized custom chips intensifies.
The Shift Back to CPU Importance
While the CPU was historically the most critical and expensive component in a server before the current AI boom, the nature of modern AI is changing the landscape. When ChatGPT launched in 2022, the first generation of AI servers typically paired up to eight GPUs to one CPU, signaling a major shift toward GPU dependency. However, the rise of agentic AI—which can operate autonomously with minimal human intervention—has renewed focus on the CPU, as this component is necessary to manage and feed data to the AI agents.
Financial markets have reacted to this shift, with CPU incumbents AMD and Intel showing strong performance in 2026, rising 128% and 149% respectively. In comparison, Nvidia’s stock had risen by 8%. Ian Buck, Nvidia’s vice president of hyperscale, stated at a recent presentation that agents have made CPUs “much more integral,” specifically noting the importance of “how fast a CPU can answer one question.”
Technical Specifications and Market Insights
Regarding market potential, Nvidia suggested that the total server CPU market could eventually reach $200 billion, a figure that contrasts with a previous Bernstein estimate of about $37 billion for the mature server CPU market in 2025. Wolfe Research, in May, predicted that the average selling price for a Vera chip would be approximately $5,000, forecasting that Nvidia would ship about 1.3 million units this year. Nvidia did not provide pricing details.
Industry analysts provided differing views on the competition. Gartner analyst Kevin Knox noted that AMD was currently the most viable option in enterprise AI server CPUs. He pointed out that while Intel reportedly holds a 66.8% market share and AMD holds about 33%, AMD is steadily gaining ground and maintaining strong relationships with large cloud service providers. Knox commented, “AMD’s done a great job building their ecosystem around their chips.”
Nvidia highlighted that Vera is the first server CPU designed from the core, rather than utilizing a pre-existing design from Arm, which reduces engineering work. The company asserts that Vera was specifically engineered to resolve bottlenecks encountered by AI agents, claiming that its chip delivers 50% better performance for AI agents compared to x86 chips, the architecture used by Intel and AMD. Nvidia explained that while previous CPUs from Intel and AMD emphasized core count, the Vera chip and its Olympus core focus instead on single-core speed. Hannah Coutand, a Vera product marketer at Nvidia, stated that the chip’s focus on high memory bandwidth, per-core speed, and low latency is designed “so that agents can return to their GPUs as quickly as possible and keep those GPUs, which are a very expensive and a highly valuable asset in the AI factory, as highly utilized as possible.”
Deployment and Technical Capabilities
Nvidia confirmed that the chip would be available standalone, in addition to being paired with the company’s GPUs. Configurations include a liquid-cooled rack containing 256 Vera chips, as well as a single server setup with two Vera chips. This combined system is marketed under the name Vera Rubin.
Technically, the CPU is noted for its power consumption, utilizing between 250 watts and 450 watts. It is also designed to accommodate a substantial amount of low-power memory, the same type found in laptops and phones, with support for up to 1.5 terabytes of memory per chip.
While some analysts, such as Karl Freund of Cambrian AI Research, suggest that Nvidia’s new CPU is intended exclusively for intensive AI tasks and not for traditional server functions like serving websites, adoption remains a question mark. Coutand noted that Vera is in an “early innings” of adoption. Although Nvidia listed Oracle among its partners, the company revealed that OpenAI intends to deploy Vera chips in large quantities starting in the current quarter. Freund observed that “The CPU is something they’ve done to kind of unhook their customers from using Intel or AMD CPUs, and they covet that revenue,” adding that Nvidia “decided to focus on a unique CPU that isn’t available in the market from anyone right now.”
