Arista Networks Launches 1.6T Platforms for AI Infrastructure

What happened

Arista Networks announced the 7060XE7 Series, a portfolio of 1.6T networking platforms designed as the foundation for rack-scale artificial intelligence systems. The platforms incorporate liquid cooling and low-power optics to help customers build AI fabrics optimised for performance and power efficiency. According to Arista's Senior VP of Cloud and AI Networking, Tyson Lamoreaux, the launch addresses the complexities of AI supersystems, where the network has evolved from a single layer to a vital component of highly integrated infrastructure. The new platforms support both scale-out and scale-up AI fabrics, meeting the extreme requirements of AI workloads that now extend to hundreds of thousands of XPUs.

Why it matters

The 7060XE7 Series marks Arista's strategic shift from providing high-performance switches to delivering comprehensive rack-scale solutions for AI infrastructure. As major technology companies are estimated to spend $750 billion on AI infrastructure this year alone, Arista's new platforms position the company to capture growing demand for specialised networking equipment. The timing is significant as the AI semiconductor market expands beyond closed-system approaches, creating opportunities for open networking ecosystems. Arista's customer concentration with large tech players like Microsoft and Meta provides some protection through switching costs, as 94% of customers report being strongly positive about Arista's products. The company's Q1 2026 results showed revenue growth of 35.1% to $2.71 billion, with raised full-year guidance of 27.7% growth, demonstrating strong market demand for its AI networking solutions.

Bigger picture

The launch reflects broader industry trends as AI workloads grow increasingly complex and power-intensive. The shift towards rack-scale solutions addresses challenges that standalone networking equipment cannot solve, including thermal management, power efficiency, and the integration of hundreds of thousands of XPUs. The networking equipment market is benefiting from the same AI infrastructure boom driving semiconductor companies, though networking firms often operate under the radar of chip makers and data center giants. The growing adoption of XPUs by large cloud providers and LLM developers is creating opportunities for open networking ecosystems, potentially reducing reliance on closed-system approaches. Memory price increases due to supply shortages have created margin pressure across the industry, affecting companies like Arista that use memory in their hardware systems. Arista's stock gained nearly 32% in Q2 2026, reflecting investor confidence in the company's positioning within the AI infrastructure buildout.

What to watch

Monitor Arista's ability to manage gross margins, which management guided to 62% to 64% for 2026, slightly down from 64.1% in 2025 due to elevated memory costs. Watch for customer adoption rates of the 7060XE7 Series and whether Arista can maintain its strong customer satisfaction metrics. The company's Q2 2026 guidance of $2.8 billion in revenue and $0.88 EPS will be important indicators of demand momentum. Pay attention to competitive dynamics as the market for AI networking solutions expands and whether Arista can benefit from the shift towards open networking ecosystems. The company's campus revenue target of $1.25 billion for the full year will signal diversification beyond cloud providers. Finally, watch for signs that AI inference workloads and XPU deployments are driving sustained demand for front-end networking infrastructure upgrades.