Skip to content
    ANET
    Earnings call· Sep 2025(Q3 FY25)

    Arista Networks, Inc. ANET

    Nov 4, 2025 Source

    Executive summary

    Arista Networks Q3 FY25 — Record Growth Driven by AI and Cloud Momentum

    Arista Networks delivered a record Q3 FY25, driven by strong demand in AI, cloud, and enterprise segments, with revenue exceeding expectations. The company is experiencing unprecedented momentum in AI build-outs, leading to increased revenue and AI-specific targets for FY26. While supply chain lead times continue to pose challenges for shipments, Arista remains confident in its demand trajectory and strategic investments in product innovation and market expansion.

    Highlights

    5
    • Achieved $2.31 billion in revenue, up 27.5% year-over-year, exceeding guidance of $2.5 billion (transcription note: guidance was $2.25B-$2.3B, actual was $2.31B, so it exceeded guidance, but the transcript states 'above our guidance of $2.5 billion' which is an ASR error).

    • Non-GAAP gross margin of 65.2% was above guidance of 64%, driven by favorable mix and inventory benefits.

    • Diluted EPS for the quarter was $0.75, up 25% from the prior year.

    • Generated approximately $1.3 billion of cash from operations, reflecting strong business model performance.

    • Raised FY26 revenue target to $10.65 billion (from $8.75B), representing 20% growth, with AI center target of $2.75 billion.

    Concerns

    5
    • International revenues decreased to 20.2% of total revenue, down from 21.8% in the prior quarter.

    • Gross margin guidance for Q4 FY25 is 62% to 63%, implying a sequential decline from 65.2% in Q3, influenced by mix towards cloud/AI titans.

    • Operating margin guidance for FY26 is 43% to 45%, lower than FY25's anticipated 48%, due to investments for strategic goals.

    • Lead times on many components range from 38 to 52 weeks, impacting ability to ship despite strong demand.

    • Inventory increased to $2.2 billion, up from $2.1 billion in the prior period, with purchase commitments totaling $7 billion, up from $5.7 billion in Q2.

    Guidance & targets

    17
    CategoryTargetConfidence
    Q4 FY25 Revenue
    $2.3 billion to $2.4 billion
    high materiality
    High
    Q4 FY25 Gross Margin
    62% to 63%
    medium materiality
    High
    Q4 FY25 Operating Margin
    approximately 47% to 48%
    medium materiality
    High
    Q4 FY25 Effective Tax Rate
    approximately 21.5%
    low materiality
    High
    Q4 FY25 Diluted Shares
    approximately 1.281 billion shares
    low materiality
    High
    FY25 Revenue Growth
    approximately 26% to 27%
    high materiality
    High
    FY25 Total Revenue
    $8.87 billion at the midpoint
    high materiality
    High
    FY25 Campus Segment Revenue
    $750 million and $800 million
    medium materiality
    High
    FY25 AI Center Target
    at least $1.5 billion
    high materiality
    High
    FY25 Gross Margin
    approximately 64%
    medium materiality
    High
    FY25 Operating Margin
    roughly 48%
    medium materiality
    High
    FY26 Revenue Growth
    approximately 20%
    high materiality
    High
    FY26 Total Revenue
    $10.65 billion
    high materiality
    High
    FY26 Campus Target
    $1.25 billion
    medium materiality
    High
    FY26 AI Center Target
    $2.75 billion
    high materiality
    High
    FY26 Gross Margin
    approximately 62% to 64%
    medium materiality
    High
    FY26 Operating Margin
    approximately 43% to 45%
    medium materiality
    High

    Segment performance

    1
    SegmentRevenueYoYQoQMargin
    International
    International revenues were $468.3 million or 20.2% of total revenue, down from 21.8% in the prior quarter.
    $468.3 million

    Operational metrics

    22
    Revenue
    $2.31 billionup 27.5% YoY
    Q3 FY25

    Arista's total revenue for the quarter, exceeding guidance.

    Software and services revenue
    18.7%
    Q3 FY25

    Software and services contribution to total revenue.

    Non-GAAP gross margin
    65.2%up from 64.6% YoY, down from 65.6% QoQ
    Q3 FY25

    Non-GAAP gross margin for the quarter, above guidance of 64%.

    Operating expenses
    $33.3 millionup from $370.6 million QoQ
    Q3 FY25

    Total operating expenses for the quarter.

    R&D spending
    $251.4 millionup from $243.3 million QoQ
    Q3 FY25

    Research and development expenses.

    Sales and marketing expense
    $109.5 millionup from $105.3 million QoQ
    Q3 FY25

    Sales and marketing expenses.

    G&A costs
    $22.4 millionup from $22 million QoQ
    Q3 FY25

    General and administrative costs.

    Operating income
    $1.12 billion
    Q3 FY25

    Operating income for the quarter.

    Other income and events
    $98.9 million
    Q3 FY25

    Favorable other income and events.

    Effective tax rate
    21.2%
    Q3 FY25

    Effective tax rate for the quarter.

    Net income
    $962.3 million
    Q3 FY25

    Net income for the quarter.

    Diluted share count
    1.277 billion shares
    Q3 FY25

    Diluted share number used for EPS calculation.

    Diluted EPS
    $0.75up 25% YoY
    Q3 FY25

    Diluted earnings per share for the quarter.

    Cash, cash equivalents and investments balance
    $10.1 billion
    Q3 FY25

    Cash, cash equivalents and investments at quarter end.

    Share repurchase program remaining authorization
    $1.4 billion
    Q3 FY25

    Remaining amount available for repurchase from the approved program.

    DSOs
    59 daysdown from 67 days QoQ
    Q3 FY25

    Days Sales Outstanding.

    Inventory turns
    1.4xflat QoQ
    Q3 FY25

    Inventory turns for the quarter.

    Inventory
    $2.2 billionup from $2.1 billion QoQ
    Q3 FY25

    Inventory balance at quarter end.

    Deferred revenue balance
    $4.7 billionup from $4.1 billion QoQ
    Q3 FY25

    Total deferred revenue balance.

    Product deferred revenue increase
    $625 millionvs last quarter
    Q3 FY25

    Increase in product deferred revenue.

    Accounts payable days
    55 daysdown from 65 days QoQ
    Q3 FY25

    Accounts payable days.

    Capital expenditures
    $30.1 million
    Q3 FY25

    Capital expenditures for the quarter.

    Industry KPIs

    8
    MetricValueDetails
    Capital return$1.4 billionUSD
    Backlog order book$7 billionUSD
    Orders backlog qualityDemand greater than ability to ship
    Ai cloud infrastructure orders$1.5 billionUSD
    Recurring software service revenue18.7%% of revenue
    Revenue mix by product customer typeAmericas: ~80%; International: 20.2%%
    Design wins product cycle transitions800-gig and 1.6-terabit
    Front end vs back end scale up vs scale across m60/40 split (approximate)

    Orderbook & backlog

    2
    Purchase commitments and inventory$7 billionQ3 FY25

    up from $5.7 billion at the end of Q2

    Reflects combination of demand for new products and lead times from key suppliers.

    Total deferred revenue balance$4.7 billionQ3 FY25

    up from $4.1 billion in Q2

    Majority is product related; can move significantly on a quarterly basis independent of underlying business drivers.

    Product announcements

    4
    ProductTypeDetails
    Etherlink portfolioroadmap
    Arista AVA (Autonomous Virtual Assist AI)launch
    Swab (Switch Aggregation Technology)launch
    Ethernet for Scale-Up Network (E.SUN) specificationlaunch

    Deals & partnerships

    2
    Oracle ExelonFormal announcement of operation with Oracle Exelon, building on a decade of partnership.decade

    This partnership builds upon a decade of collaboration with Oracle, starting with Exadata migration from InfiniBand to Ethernet for AI networks, to Rocky RDMA over converged Ethernet, and now multiplanar networking across cloud AI for on-time job completion and gigawatt scale AI data centers. Announced at Oracle AI world by Ken Duda.

    NVIDIA, AMD, Anthropic, ARMM, Broadcom, Open AI, Pure Storage, VAST DataCollaboration to build a broad and open ecosystem for the modern AI stack.

    Arista interoperates with NVIDIA and recognizes its responsibility to create a broad and open ecosystem including AMD, Anthropic, ARMM, Broadcom, Open AI, Pure Storage and VAST Data to build the modern AI stack of the 21st century. This stack includes compute, memory, storage and a solid network foundation to run training and inference models.

    Capital programs

    1
    Expanded facilities in Santa Claraunderway
    Period spend: $100 million
    Start: October 2024

    Initial construction work began in October 2024, with approximately $100 million in CapEx expected during fiscal year 2025 for this project.

    Risks & headwinds

    4
    Supply chain lead timesongoing

    38 to 52 weeks for many components (standard memory, chips, merchant silicon)

    Mitigation: Increased purchase commitments; close collaboration with large customers for early forecasts and planning.

    Gross margin pressure from customer mixQ4 FY25 and FY26

    Q4 FY25 gross margin guidance of 62-63%, down from 65.2% in Q3 FY25

    Mitigation: Careful management of mix across scale-up, scale-out, and scale-across deployments to maintain overall profitability; leveraging higher margins from enterprise customers.

    Operational expenditure increaseQ3 FY25 and beyond

    R&D up to $251.4M (10.9% of revenue), S&M up to $109.5M (4.7% of revenue)

    Mitigation: Investments are strategic, allowing for achieving strategic goals of Arista, particularly in AI and campus expansion, driven by additional headcount and acquisitions like VeloCloud.

    Variability in revenue recognition for AI projectsongoing

    Not quantified, but noted as impacting quarterly revenue

    Mitigation: Acknowledged that large AI use cases with acceptance clauses do not follow a seasonal model and 'land when they land', requiring flexibility in financial outlook.

    What to watch in Q4 FY25

    5

    AI cluster growth for fourth titan

    end of FY25 or next year
    Current3 titans crossed 100,000 GPU mark, 4th is getting there
    Target4th titan crosses 100,000 GPU mark

    Why it matters

    Indicates continued strong adoption and deployment of Arista's networking solutions for large-scale AI infrastructure by key customers.

    As you know🎣, all 4 are doing well on the 100,000 mark. 3 have already crossed it. The fourth one, I don't know if they'll cross it by end of the year or next year, but they're getting there.

    Q&A highlights

    6

    Why is there a deceleration in sequential revenue growth from 10% to 1.6% in Q4 guidance, and should investors be concerned about future growth?

    Jayshree Ullal stated there is no concern about demand, attributing variability to supply chain constraints affecting shipments rather than underlying demand. She emphasized strong demand and commitment to 20%+ growth, with the timing of revenue recognition being the main factor.

    There is no concern on our demand. I think the shipments and the revenue follows based on our supplies. So if we're able to make the shipments, then the than the revenue as you saw in Q2, when it due passed any of our guidance, right? However, there are times we can't ship everything despite the demand. And so you're accordingly seeing that. I wouldn't read too much into the quarterly variances.

    asked by Tal Liani · answered by Jayshree Ullal

    2 min read6 chapters

    Detailed Narrative

    01

    AI and Cloud Momentum

    Arista is experiencing unprecedented🌐 demand and scale in AI build-outs, with a stated goal of $1.5 billion in AI aggregate for FY25, comprising both back-end and front-end solutions. The company has committed to a new target of $2.75 billion for AI in FY26, contributing to a total revenue target of $10.65 billion. This growth is driven by momentum across cloud and AI titans, neo-cloud providers, and campus enterprise segments, as the company moves data faster across multiplanar networks.

    02

    Product Innovation and Ecosystem

    Arista continues to innovate with its Etherlink portfolio, focusing on accelerated networking for AI, bringing single-point network control for automation, security, traffic engineering, and telemetry. The company emphasizes a broad and open ecosystem, interoperating with NVIDIA, AMD, Anthropic, Broadcom, and others. Recent innovations include Arista AVA (Autonomous Virtual Assist AI) for network operations and Swab for campus stacking, leveraging a common EOS and NetDI platform across various market segments.

    03

    Leadership and Organizational Changes

    The company announced the promotion of Ken Duda to President and Chief Technology Officer, overseeing engineering and the top AI and cloud segment. Tyson Lamoreaux, an industry veteran who built the first cloud network for Amazon AWS and pioneered the first AI network for a sovereign AI company, has joined Arista to lead the cloud and AI mission and organization, reporting to Ken and Hugh. These leadership changes aim to address multiple facets of cloud and AI innovation at a system-wide level.

    04

    White Box Strategy and Market Opportunity

    Arista clarifies its role in the white box market, acknowledging that commodity white boxes are sufficient for simple use cases, while customers seek the value of Arista's 'blue boxes' for state-of-the-art hardware and troubleshooting. The 'best' option remains Arista-branded EOS platforms for ultimate superiority. The company sees an increasing total addressable market (TAM) of over $100 billion in forthcoming years, driven by the explosive AI megatrend and its 'centers of data' strategy.

    05

    Gross Margin Dynamics and Supply Chain

    While Q3 non-GAAP gross margin was strong at 65.2%, the company anticipates Q4 gross margin to be 62% to 63%, influenced by a mix shift towards cloud and AI titans, which typically have lower product margins. Management noted that lead times on many components, including standard memory, chips, and merchant silicon, range from 38 to 52 weeks, impacting the company's ability to ship everything despite strong demand. This variability in shipments, rather than demand, affects quarterly revenue recognition.

    06

    Enterprise and Campus Growth Initiatives

    Arista is aggressively investing in its enterprise go-to-market strategy, focusing on expansion into the campus segment, geographic expansion (e.g., Asia), and reaching new logos through channel investments. The VeloCloud acquisition has completed its portfolio in this area, and the company sees a significant $30 billion TAM in enterprise. This segment, alongside AI, is expected to drive substantial double-digit growth, complementing the core business which is already on large numbers.

    AI-generated summary of the company’s earnings call. Not investment advice.