Arista Is Growing Faster Than Cisco, While Cisco Brings Broader AI Scale

Arista's revenue growth and margins lead the comparison, but Cisco's $4 billion of AI orders and broader customer base make the networking contest more balanced.
Arista Networks and Cisco are both benefiting from artificial-intelligence infrastructure spending, but they occupy different positions in the market. Arista is the faster-growing specialist. Cisco is the broader incumbent with a much larger enterprise footprint.
Arista reported second-quarter revenue of $3.04 billion, its first quarter above $3 billion and a 37.7% increase from a year earlier. Non-GAAP operating margin reached 49.9%, and management guided the next quarter to about $3.3 billion. The company also raised its 2026 revenue outlook to roughly $12.6 billion, or about 40% growth.
Cisco's networking revenue was $9.79 billion, up 28%. Product orders rose 35%, or 25% excluding large hyperscale customers, and the company said AI infrastructure orders reached $4 billion. Cisco's non-GAAP operating margin of 35.9% trails Arista's, but its scale, enterprise relationships and broader security and observability portfolio provide diversification.
The technical contest is no longer simply cloud switching versus legacy networking. Arista is expanding its Etherlink platform and said it had more than 100 customers, while Cisco is using Silicon One across hyperscale and enterprise environments. Cisco reported three design wins that span those markets.
Arista's concentration is both an advantage and a risk. A small group of large cloud customers can drive rapid growth, but purchasing shifts by any one of them can affect results. Management also expects supply constraints to remain relevant into 2028. Cisco is less concentrated, yet its gross margin declined to 66.3% from 68.4% as hardware mix increased.
For investors, the central question is how much of today's AI buildout becomes durable network demand. Arista currently converts that demand into faster growth and higher margins. Cisco offers lower execution concentration and a wider installed base, but must prove it can keep winning the highest-performance AI workloads. The companies can both gain as network traffic rises, even if the market continues to assign them very different valuation multiples.



