Skip to content
    AVGO
    Earnings call· Jul 2026(Q3 FY26)

    Broadcom Q3 FY26 earnings call AVGO

    Sep 2, 2026 Source

    Executive summary

    Broadcom Q3 FY26 — AI Semiconductor Revenue Triples, Long-Term Outlook Raised

    Broadcom delivered an exceptional quarter, driven by robust demand for its AI semiconductor solutions, particularly custom accelerators (XPUs) for hyperscalers and LLM customers. The company provided significantly raised long-term AI revenue guidance through FY28, underpinned by strategic partnerships and an expanding product portfolio in AI networking. Management emphasized strong operating leverage despite gross margin compression from product mix, and is actively addressing supply chain and infrastructure deployment challenges.

    Highlights

    5

    • Consolidated revenue reached a record $29.6 billion, up 86% year-on-year.

    • AI semiconductor revenue grew 221% year-on-year to $16.7 billion in Q3, and is expected to reach $21.7 billion in Q4, up 236% year-on-year.

    • Operating income grew 92% year-on-year, with operating margin at a record 68% of revenue.

    • Non-GAAP EPS of $3.32 was up 96% year-on-year.

    • Long-term AI revenue guidance raised to $115 billion for FY27 and $230 billion for FY28, with a target to exceed $30 in EPS by FY28.

    Concerns

    3

    • Consolidated gross margin declined 210 basis points sequentially to 75% in Q3, and is expected to be 73% in Q4, due to the increasing mix of XPUs with higher memory content.

    • Supply chain bottlenecks, including leading-edge silicon, substrates, HBM memory, and system memory, remain a constant challenge for meeting demand.

    • Constraints related to land, power, and data center shell availability impact the timely deployment of compute capacity.

    Guidance & targets

    13
    CategoryTargetConfidence
    Consolidated Revenue
    $34.8 billion
    high materiality
    High
    AI Semiconductor Revenue
    $21.7 billion
    high materiality
    High
    Fiscal Year AI Revenue
    $58 billion
    high materiality
    High
    AI Semiconductor Revenue
    approximately $115 billion
    high materiality
    High
    AI Semiconductor Revenue
    $230 billion
    high materiality
    High
    Non-AI Semiconductor Revenue
    approximately $4.3 billion
    medium materiality
    High
    Infrastructure Software Revenue
    approximately $8.7 billion
    medium materiality
    High
    Consolidated Operating Margin
    approximately 66%
    high materiality
    High
    Consolidated Gross Margin
    approximately 73%
    medium materiality
    High
    Non-GAAP Tax Rate
    approximately 16%
    low materiality
    High
    Non-GAAP Diluted Share Count
    approximately 4.94 billion shares
    low materiality
    High
    Capital Expenditures
    $1.4 billion
    medium materiality
    High
    EPS
    exceed $30
    high materiality
    High

    Segment performance

    2
    SegmentRevenueYoYQoQMargin
    Semiconductor Solutions
    Revenue was a record, representing 70% of total revenue. AI semiconductor revenue increased from 49% of total revenue in Q2. Operating margin increased 440 basis points year-on-year due to revenue growth outpacing operating expenses.
    AI semiconductor revenue: $16.7 billionAI semiconductor revenue as % of total revenue: 56%Gross margin: 76%Operating expenses: $1.2 billionOperating expenses as % of segment revenue: 6%Operating expenses growth year-on-year: 22%
    		$20.8 billion127%—61% operating margin
    Infrastructure Software
    Revenue represented 30% of total revenue. Software operating margin increased 650 basis points year-on-year.
    Gross margin: 94%Operating expenses: over $900 millionARR growth year-on-year: 15%
    		$8.8 billion29%—84% operating margin

    Operational metrics

    17
    Non-GAAP EPS
    $3.32up 96% year-on-year
    Q3 FY26

    Reported for the quarter.

    Consolidated Operating Margin
    67.9%up 240 basis points year-on-year
    Q3 FY26

    Record operating margin, reflecting strong operating leverage.

    Consolidated Gross Margin
    75%down 210 basis points sequentially
    Q3 FY26

    Better than guidance of 74%, but declined due to AI semiconductor revenue being a greater proportion of total revenue mix.

    Cash and investments balance
    $24 billionup $4.3 billion sequentially
    Q3 FY26

    Cash balance at the end of the third quarter.

    Long-term debt paid down
    $5.6 billion
    Q3 FY26

    Amount of long-term debt paid down in the quarter.

    Long-term debt paid down (subsequent)
    $1.5 billion
    post Q3 FY26

    Additional senior notes paid down upon maturity subsequent to quarter end.

    Gross principal fixed rate debt
    $59.6 billion
    Q3 FY26

    Weighted average coupon rate and years to maturity for the gross principal fixed rate debt.

    Free cash flow margin
    46%
    Q3 FY26

    Free cash flow as a percentage of revenue.

    Capital expenditures
    $532 million
    Q3 FY26

    Amount spent on capital expenditures in the quarter.

    Dividends paid
    $3.1 billion
    Q3 FY26

    Cash dividends paid to stockholders based on a quarterly common stock dividend of $0.65 per share.

    AI semiconductor revenue growth
    54%sequentially
    Q3 FY26

    Sequential growth of AI semiconductor revenue.

    XPU shipments growth
    over 3.5xyear-on-year
    Q3 FY26

    Growth in XPU shipments, representing 73% of AI revenue.

    AI networking revenue growth
    over 2.5xyear-on-year
    Q3 FY26

    Growth in AI networking revenue.

    Non-AI semiconductor revenue
    $4.2 billionup 5% year-on-year, flat sequentially
    Q3 FY26

    Revenue for non-AI semiconductors, partially offset by decline in wireless.

    Infrastructure software ARR growth
    15%year-on-year
    Q3 FY26

    Annual Recurring Revenue growth for infrastructure software.

    Inventory
    $4.5 billion
    Q3 FY26

    Inventory balance to support strong semiconductor demand.

    Dollars per gigawatt (Broadcom content)
    $20 billion to $30 billionrelatively stable
    future

    Expected range for Broadcom's content per gigawatt, remaining stable despite increasing chip performance due to higher power per chip.

    Industry KPIs

    3
    MetricValueDetails
    Ai data center revenue$16.7 billionUSD
    Inventory channel inventory$4.5 billionUSD
    Node platform ramp scheduleIronwood TPU v7, TPU version AI, Jalapeno, MTIA accelerator

    Product announcements

    6
    ProductTypeDetails
    Ironwood TPU v7milestone
    TPU version AIlaunch
    TPU version 8imilestone
    Jalapenolaunch
    Tomahawk 7launch
    VMware Private AI Cloudlaunch

    Deals & partnerships

    3
    GoogleLong-term agreement to develop and supply future generations of TPUs and AI networking.multi tens of billions of dollars annuallyseveral years

    Partnership strengthened by a long-term agreement for successive generations of TPUs and AI networking, leveraging Broadcom's IP portfolio and fast time to market.

    Anthropic and OpenAIAI XPV platform in partnership with Apollo and Blackstone to enable compute infrastructure.

    Platform established with Apollo and Blackstone to deploy more than 20 gigawatts of compute infrastructure for OpenAI and Anthropic by the end of 2028. Broadcom may provide modest residual value guarantees.

    MetaPartnership to deliver multiple generations of MTIA XPUs.through 2027

    Delivering 3 generations of MTIA accelerators to Meta between now and the end of 2027.

    Capital programs

    3
    AI XPV platformunderwaymore than 20 gigawatts of compute infrastructure
    Spent to date: $35 billion (first tranche)
    Funding: Apollo and Blackstone (third-party financial partners)
    Start: June

    Benefit: 20+ gigawatts for OpenAI and Anthropic

    Established in partnership with Apollo and Blackstone to enable compute infrastructure for OpenAI and Anthropic. The first $35 billion tranche closed in June for Anthropic's 1 gigawatt deployment. Future tranches will have unique features tailored to specific lab and investor needs. Broadcom may provide modest residual value guarantees.

    Singapore Substrate Fabunderway

    Benefit: Substrate capacity

    Broadcom is building its own substrate capacity at scale in its factory in Singapore, starting deployment in fiscal '27, to address key supply bottlenecks.

    EML, CW, and VCSEL factories expansionunderway

    Benefit: More than tripling capacity year-on-year for EML, CW, and VCSEL lasers and Indium phosphide factories.

    Investing in factories for EML, CW, and VCSEL lasers and Indium phosphide, more than tripling capacity year-on-year, with significant increases planned for the next two years to meet demand for optical interconnects.

    Risks & headwinds

    3
    Gross margin dilution from AI product mixQ3 FY26, Q4 FY26

    Consolidated gross margin declined 210 basis points sequentially to 75% in Q3, expected to be 73% in Q4 (down from 78% a year ago).

    Mitigation: Focus on strong operating leverage, as revenue growth outpaces operating expenses, sustaining operating margin.

    Supply chain bottlenecks for AI semiconductorsOngoing, particularly for FY27 and FY28 growth targets.

    Impacts leading-edge silicon, substrates, HBM memory, and system memory.

    Mitigation: Secured supply for current outlooks; building own substrate capacity in Singapore (deploying FY27); expanding EML, CW, VCSEL factories; close collaboration with 6 key customers to manage multi-dimensional problem.

    Deployment constraints (Land, Power, Shell)Ongoing, long lead time for construction projects.

    Dictates specific timing of when compute capacity gets deployed and becomes available.

    Mitigation: Engaged with customers on their LPS readiness and reflecting this in forecast outlooks; collaborative effort to ensure timely deployment.

    What to watch in Q4 FY26

    5

    Q4 AI Semiconductor Revenue

    Q4 FY26
    Current$16.7 billion (Q3 FY26)
    Target$21.7 billion

    Why it matters

    This is a key indicator of the continued acceleration of AI demand and Broadcom's execution on its custom accelerator and networking solutions.

    And together, we expect these deployments to drive our Q4 AI revenue to $21.7 billion, which is up 236% year-on-year.

    Q&A highlights

    5

    Regarding the doubling of AI business next year and demand potentially exceeding supply, what are the supply bottlenecks, and what variables could increase the number if resolved?

    Hock Tan stated that the company is careful and conservative with its outlook, acknowledging that demand could be significantly higher. He mentioned that the company has secured supply for the $115 billion FY27 and $230 billion FY28 outlooks, but deployment timing (getting chips deployed on time) is a key consideration. He also noted that the Singapore fab will start deploying for substrates in FY27 to address bottlenecks.

    “But certainly, our customers want us to ship more. But we have secured supply, and we think it's the right number to put it to $115 billion.”

    asked by Joseph Moore · answered by Hock Tan

    2 min read6 chapters

    Detailed Narrative

    01

    AI Semiconductor Growth and Customer Adoption

    Broadcom's Q3 AI semiconductor revenue surged 221% year-on-year to $16.7 billion, driven by custom accelerators (XPUs) for six key customers. The company shipped Ironwood TPU v7 to Anthropic and Google, and began production of the next-gen TPU version AI for Google. Jalapeno, OpenAI's first-gen custom accelerator, also shipped, outperforming Grace Blackwell GPU for inference workloads. XPU shipments were up over 3.5x year-on-year, representing 73% of AI revenue.

    02

    Long-Term AI Revenue Outlook and XPU Deployments

    Broadcom significantly raised its AI revenue guidance, projecting $115 billion for FY27 and $230 billion for FY28, doubling year-over-year. This growth is supported by long-term agreements, including a multi-tens of billions of dollars annual TPU commitment with Google. Anthropic is expected to deploy 1GW of Ironwood in 2026, 5GW of TPU V8i in 2027, and an incremental 10GW in 2028. OpenAI plans 1.3GW of Jalapeno in 2027 and over 5GW in 2028, with Meta deploying 3GW of MTIA accelerators through 2028.

    03

    AI Networking Leadership and Innovation

    The company maintains its leadership in AI networking, with AI networking revenue expected to grow as fast as XPUs. Broadcom was first to market with its 100 terabit Tomahawk 6 Ethernet switch and has taped out Tomahawk 7, the industry's first 200 terabit per second switch. It also leads in PCI Express switching and is expanding capacity in optical DSPs, EMLs, VCSELs, and CW lasers for optical interconnects, providing a broad and leading-edge AI portfolio.

    04

    Infrastructure Software and Private AI Cloud

    Infrastructure Software revenue grew 29% year-on-year to $8.8 billion in Q3, with 15% ARR growth. Broadcom announced VMware Private AI Cloud, offering enterprises a secure and cost-effective platform for AI alongside existing applications. This initiative aims to repatriate workloads from public to private clouds, leveraging VCF (VMware Cloud Foundation) to improve infrastructure economics and open new opportunities for the software business.

    05

    Capital Allocation and XPV Platform

    Broadcom ended Q3 with $24 billion in cash and paid down $5.6 billion of long-term debt. The company established the AI XPV platform in partnership with Apollo and Blackstone to enable over 20 gigawatts of compute infrastructure for OpenAI and Anthropic by the end of 2028. The first $35 billion tranche for Anthropic's 1GW deployment is underway, with future tranches tailored to specific lab and investor needs, leveraging third-party financing with modest residual value guarantees.

    06

    Gross Margin Dynamics and Operating Leverage

    Consolidated gross margin was 75% in Q3, down 210 basis points sequentially, and is guided to 73% for Q4. This decline is attributed to the increasing proportion of AI semiconductor revenue, particularly XPUs with higher memory content. Despite this, operating margin reached a record 67.9% in Q3, up 240 basis points year-on-year, demonstrating strong operating leverage as revenue growth significantly outpaces operating expenses.

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