Arista Networks (ANET) Morgan Stanley Technology, Media & Telecom Conference 2026 summary

Evolving value proposition and AI-driven architecture

• Core value proposition remains mission-critical networking with foundational software, now enhanced by AI-driven architectures and all-Ethernet AI spine and leaf designs.

• Uniform network infrastructure supports diverse environments: data center, campus, WAN, and AI centers, addressing heterogeneity in models and accelerators.

• Differentiation comes from purpose-built platforms, leveraging merchant silicon with advanced software for reliability and flexibility.

Market trends, use cases, and power constraints

• Power availability is the main bottleneck for new data center builds, with demand shifting to hundreds of megawatts or gigawatts.

• Three main AI networking use cases: scale up (within rack), scale out (across racks), and scale across (between data centers), each with unique power and connectivity challenges.

• Distributed AI centers are emerging due to power limitations at single sites.

Competitive landscape and differentiation

• Product differentiation is driven by a robust platform, ecosystem partnerships, and the ability to support heterogeneous AI accelerators.

• The market is expected to remain multi-model and multi-protocol, with no single AI model or accelerator dominating.