Arista Networks Targets $12.6B as AI, Campus and Cloud Networking Growth Accelerates

Key Points

• Arista is targeting $12.6 billion in 2026 revenue, implying approximately 40% growth, while continuing to invest across AI, data-center switching, software, campus networking and routing.
• Improved supply visibility led Arista to raise annual guidance by more than $1 billion, while multiyear purchase commitments nearly tripled to $9.7 billion. The company maintained its 2026 gross-margin outlook of 62% to 64% despite ongoing component constraints.
• AI networking and campus expansion are major growth drivers: AI revenue is targeted at $3.5 billion, with scale-across networking representing about 30% of that figure, while the campus revenue target rose above $1.25 billion.

Arista Networks NYSE: ANET executives said the company is investing across its networking portfolio rather than making tradeoffs between AI and data-center switching, software, campus networking and routing, as it works toward its 2026 revenue outlook of $12.6 billion.

Speaking at the Goldman Sachs Communacopia and Technology Conference, Chief Financial Officer Chantelle Breithaupt said the outlook implies roughly 40% growth and provides substantial absolute-dollar capacity for research and development. Arista generally targets R&D spending of 8% to 10% of revenue, she said.

President and Chief Technology Officer Ken Duda said the company’s work with sophisticated hyperscale customers on new technologies and customized systems supports its broader portfolio. Investments in hardware and software for large operators can flow into specialty cloud providers and enterprise deployments, he said.

Supply Chain, Demand Visibility and Margins

Breithaupt said Arista raised its annual guidance by more than $1 billion after becoming more comfortable with supply availability and gaining greater order visibility. The company secured purchase commitments, received support from suppliers and had visibility into two quarters of purchase orders by the August timeframe, she said.

While supply conditions have improved, Breithaupt cautioned that the industry is not fully beyond component constraints. Availability can be affected by a range of items, from major components such as chips and memory to peripheral items including printed circuit boards and power cables.

Arista’s multiyear purchase commitments nearly tripled to $9.7 billion, according to the discussion. Breithaupt characterized the increase as a demand signal rather than solely the result of component-price inflation. Chips have lead times of roughly 52 weeks, she said, prompting the company to make purchasing decisions well ahead of the order visibility it has in hand.

The company is comfortable with that approach because its portfolio is relatively flexible and components can be used across products and customers, Breithaupt said. Duda added that common switch models and components serve multiple customer types and use cases, helping reduce potential inventory-obsolescence risk.

Arista maintained its 2026 gross-margin outlook of 62% to 64%. Breithaupt said customer mix remains a primary variable, while tariff refunds are expected to contribute about 30 basis points for the year. The company also implemented a targeted price increase for products with meaningful exposure to memory and other component-cost inflation. The increase was intended to offset inflation rather than expand margins, she said.

EOS and Open Networking

Duda said Arista continues to view its Extensible Operating System, or EOS, as a major differentiator. He said open-source network operating systems require extensive technical effort to assemble, validate and deploy across specific hardware platforms and customer use cases.

In hyperscale environments, open-source approaches are primarily used to support multisourcing and reduce dependence on a single software supplier, Duda said. He said Arista retains an advantage in demanding routing applications that require fast convergence, large routing tables, policy controls, tunnel encapsulation capabilities and operation within hardware-memory limits.

For cloud providers and enterprises, Duda said EOS, along with Arista’s CloudVision management platform, offers consistent operations across data centers, campuses, wide-area networks and cloud deployments. The same operating-system code runs across those environments, he said.

AI Networking Opportunities

Executives highlighted AI infrastructure as a significant growth driver, particularly among NeoCloud providers. Duda said these customers seek best-of-breed technologies to optimize their full technology stacks and reduce token costs, rather than relying on a single-vendor system.

Arista’s offerings include networking capabilities for scale-out AI clusters, as well as CloudVision visibility into both network conditions and AI-server metrics such as flow control, congestion and retransmissions, Duda said. Breithaupt said the company remains selective in its commercial arrangements with NeoCloud customers, using measures such as prepayment where appropriate because not all emerging providers have equivalent financial backing.

Duda defined scale-across networking as the interconnection of AI clusters split across data centers. Unlike scale-out networking within a data center, scale-across applications require greater routing complexity and deeper buffering to account for the longer round-trip distance between geographically distributed clusters, he said.

Power, cooling and data-center-space constraints are driving customers to distribute GPU deployments across locations, creating demand for scale-across architectures and Arista’s 7800R Series platforms, according to Duda. Breithaupt said scale-across represented about 30% of Arista’s stated $3.5 billion AI revenue target for the year.

The company also sees a longer-term opportunity in scale-up networking, which connects accelerators within a rack or enclosure. Duda said Arista has limited share in that market today but expects Ethernet-based standardization, including ESUN, to create an opening. He expects activity to begin ramping in the latter part of 2027, with volume potentially arriving in early 2028.

Campus Growth and Capital Priorities

Arista raised its campus revenue target to more than $1.25 billion for the year. Duda said the company initially won campus business from existing data-center customers, but is now seeing a growing share of new campus customers seeking alternatives to incumbent vendors.

The company cited its Cognitive Campus strategy, which includes always-on operations, zero-touch capabilities and zero-trust networking. Duda said Arista’s in-service software upgrades have become a differentiator for campuses supporting 24-hour operations such as hospitals, manufacturing facilities, logistics centers and media operations.

Breithaupt said Arista has roughly 5% market share in campus networking and views the segment as a high-volume, lower-dollar growth opportunity over multiple years.

On capital allocation, Breithaupt said working capital is the company’s first priority as it navigates what she described as step-function growth. Share repurchases remain opportunistic, while Arista continues to evaluate acquisition opportunities but has not identified a target that meets both its technology and cultural criteria.

Looking ahead, Duda said he is particularly interested in applying AI internally and within Arista’s products to build network engineering and operations assistants. Such tools could improve customer efficiency, uptime and operational management, he said.

About Arista Networks (NYSE:ANET)

Arista Networks, Inc designs, develops and sells cloud networking solutions for large data centers, cloud service providers, internet companies, enterprises and other organizations. Its portfolio includes high-performance Ethernet switches, routers and wireless networking products used to connect servers, storage systems, users and applications across data center and campus environments.

The company's software offerings include Arista EOS, its Linux-based network operating system, and CloudVision, a cloud-based platform for network management, automation, monitoring and analytics.

This instant news alert was generated by narrative science technology and financial data from MarketBeat in order to provide readers with the fastest reporting and unbiased coverage. Please send any questions or comments about this story to contact@marketbeat.com.