Nvidia (NVDA) Stock; Dips on Arista Deployments Favoring AMD

TLDRs;

• Nvidia stock dips nearly 3% as Arista shifts 20-25% of deployments to AMD.
• AMD gains market share as Arista supports multi-vendor AI data center setups.
• Nvidia’s Spectrum-X Ethernet platform competes with Arista’s growing open ecosystem.
• AI infrastructure battle intensifies, signaling diversification among large-scale operators.

Shares of Nvidia (NVDA) declined nearly 3% on February 13, reflecting investor caution after Arista Networks revealed a noticeable pivot toward AMD hardware in its AI data center deployments. This shift marks a significant change from last year when nearly all Arista deployments relied on Nvidia technology. Meanwhile, AMD (AMD) shares saw a modest uptick of around 1%, buoyed by the news of increased adoption.

Arista, known for its high-performance Ethernet switching technology that underpins AI server clusters, plays a substantial role in the AI infrastructure ecosystem. Although Nvidia is now producing its own networking solutions, including the Spectrum-X Ethernet platform launched in 2023, the market reaction indicates that Arista’s deployment choices still carry weight among institutional investors.

Arista Expands AMD Footprint

According to Arista’s CEO, approximately 20-25% of the company’s deployments now favor AMD, signaling a diversification strategy in its AI infrastructure offerings. The move highlights a shift toward multi-vendor solutions, reducing the risk of dependency on a single supplier and allowing customers to optimize hardware configurations for performance and cost.

NVIDIA Corporation, NVDA

Arista’s AI networking unit is a growing business, having generated $1.5 billion in revenue in 2025, with projections to rise to $3.25 billion in 2026. This growth stems from close partnerships with major AI data center operators, whose clusters collectively operate hundreds of thousands of GPUs. Such scale ensures that Arista’s choices can meaningfully impact market sentiment, as reflected in the recent Nvidia stock movement.

The Platform Battle in AI Data Centers

The current market developments reflect a broader competition over AI data center infrastructure design. Nvidia is advocating a vertically integrated “full stack” approach, coupling its GPUs with the Spectrum-X networking platform for optimized performance across the compute-to-network continuum. Major players such as Meta and Oracle are already adopting Nvidia’s integrated stack for certain AI workloads.

In contrast, Arista is promoting an open ecosystem strategy. Its AI infrastructure stack now incorporates AMD, Anthropic, Arm, Broadcom, OpenAI, Pure Storage, and VAST Data, offering customers flexibility and vendor diversity. The partial adoption of AMD hardware indicates that some AI operators prefer a combination of suppliers to avoid vendor lock-in and enable adaptable, cost-effective solutions.

Implications for Nvidia and AI Investors

The shift by Arista serves as a reminder that market leadership in AI GPUs does not guarantee dominance in networking. Nvidia’s foray into its own Ethernet solutions faces competition from established providers like Arista, which are increasingly offering multi-vendor options aligned with customer demand for flexibility.

Investors and industry observers will be closely watching how Nvidia responds to this diversification trend. While its GPU technology remains critical for AI workloads, the growing presence of AMD in Arista’s deployments may influence future hardware selection decisions across the AI data center market.

As the AI infrastructure landscape evolves, companies like Nvidia and AMD are no longer just competing on chips, they are contending for the entire ecosystem, from servers and networking hardware to integrated software stacks. The Arista development underscores that strategic partnerships and ecosystem openness could play as much a role in investor sentiment as raw technology performance.

The post Nvidia (NVDA) Stock; Dips on Arista Deployments Favoring AMD appeared first on CoinCentral.