AVGO · Q1 2026 Earnings

Bullish

Broadcom

Reported March 4, 2026

30-second summary

30-second take. Broadcom delivered Q1 revenue of $19.31B (+29.4% YoY), beating consensus of $19.11B by 1.1% and the $19.1B prior guide, with AI semis at $8.4B (+106% YoY) beating the $8.2B guide by $200M. Q2 FY2026 is guided to $22.0B (+47% YoY off the $15.0B Q2 FY2025 base) with AI semis stepping to $10.7B (+140% YoY) — a $2.3B sequential dollar increase, the largest single-quarter step-up Broadcom has ever guided. The signal that matters: management put a $100B+ AI silicon revenue line of sight on the table for 2027, named OpenAI as the sixth XPU customer with one-gigawatt 2027 deployment, and disclosed that supply chain is fully secured through 2028 — converting what was a qualitative multi-year ramp narrative into a quantified, supply-locked commitment.

Headline numbers

EPS

Q1 FY2026

$2.05

+1.5% vs est.

Revenue

Q1 FY2026

$19.31B

+29.4% YoY

+1.1% vs est.

Gross margin

Q1 FY2026

68.1%

Free cash flow

Q1 FY2026

$8.01B

Operating margin

Q1 FY2026

44.3%

Key financials

Q1 FY2026

Metric	Q1 FY2026	YoY	Q4 FY2025	QoQ
Revenue	$19.31B	+29.4%	$18.02B	+7.2%
EPS	$2.05	—	$1.95	+5.1%
Gross margin	68.1%	—	68.0%	+12bps
Operating margin	44.3%	—	41.7%	+263bps
Free cash flow	$8.01B	—	$7.47B	+7.3%

Guidance

Guidance is issued one quarter forward. The Prior-guide column references the guide issued last quarter for the period just reported; the New-guide column is for next quarter.

Actuals vs prior guidance

Metric	Period	Prior guide	Actual	Δ	Result
Revenue	Q1 FY2026	approximately $19.1 billion	$19.311 billion	in-line	Met
Revenue YoY growth	Q1 FY2026	28%	29.4%	+1.4 percentage points above guide	Beat
Adjusted EBITDA margin	Q1 FY2026	67% of projected revenue	68%	in-line	Met
AI Semiconductor Revenue	Q1 FY2026	$8.2 billion	$8.4 billion	+$0.2 billion above guide	Beat
Semiconductor Solutions Revenue	Q1 FY2026	$12.3 billion (implied: $8.2B AI + $4.1B non-AI)	$12.515 billion	+$0.2 billion above guide	Beat
Infrastructure Software Revenue	Q1 FY2026	approximately $6.8 billion	$6.796 billion	in-line	Met
Gross Margin	Q1 FY2026	down approximately 100 basis points sequentially	68.1%	Actual margin declined less than 100bp sequentially; sequential impact better than guided	Beat

New guidance

Metric	Period	Guide	YoY
Revenue	Q2 FY2026	approximately $22.0 billion	+47% YoY
Revenue YoY growth	Q2 FY2026	47%	—
Adjusted EBITDA margin	Q2 FY2026	approximately 68% of projected revenue	—
AI Semiconductor Revenue	Q2 FY2026	$10.7 billion	—
Semiconductor Revenue	Q2 FY2026	$14.8 billion, up 76% YoY	+76%

Segment performance

Q1 FY2026

Segment	Q1 FY2026	YoY
Semiconductor Solutions	$12.515B	+52.4%
Infrastructure Software	$6.796B	+1.4%

Capacity & utilization

Q1 FY2026

Segment	Q1 FY2026
AI Revenue	$8.4B
AI Revenue Growth YoY	106%
Q2 AI Revenue Guidance	$10.7B
Capital Expenditures	$250M

Profitability

Q1 FY2026

Segment	Q1 FY2026
Adjusted EBITDA	$13.128B
Adjusted EBITDA Margin	68%
Free Cash Flow Margin	41%
Operating Cash Flow	$8.26B

Management tone

Narrative arc: prior quarters built from initial XPU customer disclosures to a five-customer roster with $73B backlog quantification → Q1 FY2026 sixth customer (OpenAI) + $100B+ 2027 line of sight + supply chain locked through 2028.

Management put a 2027 number on the table for the first time: "we have line of sight to achieve AI revenue from chips just chips in excess of $100 billion in 2027 we have also secured the supply chain required to achieve this." The structural shift is the supply-chain pairing — management is no longer guiding a demand-side aspiration but a delivered-volume commitment. That is unusually specific forward language for a semiconductor management team that has historically held tight quarterly discipline. The Q&A with Stacy Rasgon validated a ~10 gigawatt / ~$20B per gigawatt customer-by-customer build-up methodology that management essentially endorsed as "the right way to look at it" — putting the buy-side analytical framework on the record as compatible with the headline.

The sixth XPU customer was named explicitly as OpenAI: "we expect OpenAI to deploy in volume their first generation XPU in 2027 and over one gigawatt of compute capacity." The disclosure moved from anonymized prospects to fully named with gigawatt-scale deployment specifics.

Management closed the loop on supply: "we have fully secured capacity of these components for 26 through 28" covering leading-edge wafers, HBM, and substrates. In an industry where supply is the binding constraint on the AI buildout, Broadcom is now disclosing a multi-year secured position — that converts the $100B+ 2027 number from forecast to commitment. It also raises the bar for what counts as a credible competitive threat from customer-owned tooling, which management addressed directly in Q&A.

Infrastructure Software framing turned proactive: "As the permanent abstraction layer between AI software and physical chips, silicon, VCF cannot be disintermediated or replaced." The "non-disruptible essential layer" language reframes VMware from a legacy revenue annuity to an AI-era beneficiary — though the +1.4% YoY print this quarter undercuts the rhetoric. Management is asking the buy-side to underwrite a multi-year reacceleration that the current quarter does not yet show.

Networking framing escalated. Networking is guided to 40% of AI revenue in Q2 with explicit share gain language and a 2027 Tomahawk 7 roadmap. The Q&A confirmed networking is currently growing faster than XPU growth — meaning the long-debated "networking declines as XPU mix rises" investor model is wrong. Both lines are scaling together.

Recurring themes management leaned on this quarter:

Custom AI accelerator XPU dominance with six named customers and multi-gigawatt 2027-2028 deploymentsSupply chain as secured competitive advantage through 2028 enabling durabilityInfrastructure software as AI-enabled, non-disruptible essential layerAI networking acceleration capturing share in both scale-out and scale-up architecturesRecord profitability and operating leverage with 68% adjusted EBITDA marginsLine of sight to $100B+ AI semiconductor revenue in 2027

Risks management surfaced:

Geographic mix of income and global minimum tax impact on tax ratesPotential share repurchase execution (subject to authorization and market conditions)Wireless seasonal decline offsetting strength in other non-AI segmentsConstrained capacity in leading edge wafers and HBM (mitigated by secured supply agreements)

Q&A highlights

Blaine Curtis · Jefferies

Clarification on >$100B AI chips revenue forecast (ASICs vs networking vs RAC) and whether hyperscalers' ROI requirements present a risk to demand outlook for 2027

Hawk clarified that >$100B forecast is primarily silicon-based (XPUs, switch chips, DSPs). Addressed hyperscaler ROI concerns by noting customers are building custom accelerators and networking clusters; demand for both training and inference compute capacity is very strong, particularly inference for productization and monetization of LLMs.

>$100B revenue forecast in 2027 based on silicon chipsSix primary customers (not five)Strong demand for both training and inference compute capacityCustomers building custom accelerators and custom networking cluster architecture

Harlan Sir · JP Morgan

Assessment of customer-owned tooling (COT) competitive threat from hyperscalers' internal XPU/TPU design efforts; likelihood of material share loss and how Broadcom maintains 12-18 month technology lead

Hawk provided detailed response that COT initiatives face massive challenges (technology, design complexity, advanced packaging, clustering expertise). Emphasized Broadcom's 20+ year silicon expertise and unique ability to achieve high-volume production quickly. Stated competition from COT unlikely for many years; Broadcom viewed as essential strategic partner due to performance requirements and execution capability.

COT initiatives 2x less performant than Broadcom's current generationBroadcom 12-18 months ahead in performance, complexity, and IP20+ years of silicon design and manufacturing experienceCompetitors (including Nvidia) continuously improving; XPU makers must match or exceed competing LLM platforms

Ross Seymour · Deutsche Bank

Why networking revenue now represents ~40% of AI revenue; expected mix in $100B+ forecast; leadership position in scale-out vs scale-up and cross-optimization opportunity with XPU designs

Hawk attributed networking growth to new GPU/XPU generations requiring 200G bandwidth; Tomahawk 6 switch (only 100 terabit/sec option) and 1.6 terabit optical transceiver DSPs (only player at this speed) driving demand. Expects networking to stabilize at 33-40% of AI revenue mix. Announced Tomahawk 7 (2X performance) launching in 2027; networking growth outpacing XPU growth currently.

Networking currently 40% of AI revenueTomahawk 6: only 100 terabit/sec switch available; new generation Tomahawk 7 with 2X performance launching in 20271.6 terabit optical transceiver DSP: only player at this speed200G bandwidth standard for new GPU/XPU generations

Stacey Raskon · Bernstein

Validation of >$100B forecast by reverse-engineering customer gigawatt commitments and content per gigawatt; specific breakdown of customer capacity additions

Hawk acknowledged Stacey's gigawatt-based analysis as correct approach ('right way to look at it'). Confirmed approximately 10 gigawatts of deployment expected in 2027. Acknowledged content per gigawatt varies by customer but roughly aligns with Stacey's ~$20B per gigawatt estimate. Did not dispute underlying customer assumptions (Anthropic 3GW, OpenAI 1GW, Meta 2GW+, Google 3GW+).

~10 gigawatts of customer AI chip deployments expected in 2027Content per gigawatt varies by customer but ranges around $20B per gigawattSix total customers driving the forecastAnalysis validates substantially >$100B revenue guidance

Vivek Arya · Bank of America Securities

Breakdown of Anthropic's $20B spending allocation between chips vs racks; distinction between chip revenue and rack revenue in $100B forecast; how Broadcom maintains visibility and share given fragmented multi-supplier customer environment

Hawk declined to break down Anthropic chips vs racks specifically ('rather not answer that'). Emphasized all six customers pursue XPU as strategic necessity (not optionality), citing custom accelerators like Meta's MTIA as existentially critical to competitive positioning. Distinguished XPU commitments as long-term strategic with clear visibility vs transactional GPU/cloud alternatives. Stated visibility increases as customers mature in XPU development journey.

Six customers; all pursuing custom silicon as strategic imperativeXPU programs are multi-year, non-negotiable strategic roadmapsGPU/cloud alternatives characterized as transactional and optionalBroadcom has clear multi-year visibility into customer roadmaps

Answers to last quarter's watch list

Q1 FY2026 AI semiconductor revenue vs. the $8.2B guide (+100% YoY). AI semis came in at $8.4B (+106% YoY), beating the doubling guide by $200M, and management converted the FY2026-evasion into a 2027 line-of-sight figure of $100B+ from silicon alone — well beyond what any sell-side model contemplated. This is the most aggressive forward quantification Broadcom has issued.

Resolved positively

Q1 FY2026 gross margin landing vs. the down ~100bps QoQ guide. Q1 consolidated non-GAAP gross margin came in at 77%, better than the down-~100bps QoQ guide implied, and Q2 is now guided flat sequentially at 77%. The system-sales margin compression is more gradual than telegraphed. Adjusted EBITDA margin held at 68%.

Resolved positively

Sixth XPU customer qualification. OpenAI was named explicitly as the sixth qualified customer with first-generation XPU production deployment in 2027 and over one gigawatt of compute capacity.

Resolved positively

AI backlog growth from $73B. Backlog was not re-quantified in this print — management pivoted from backlog framing to a 2027 silicon revenue line-of-sight figure of $100B+, with Q&A walk-through of customer gigawatt commitments. The disclosure framework shifted from backlog to forward revenue with supply chain secured through 2028, which is arguably stronger but breaks comparability. Status: Not resolved (disclosure framework changed).

Non-AI semiconductor Q1 print vs. the $4.1B flat-YoY guide. Non-AI semis came in at approximately $4.1B (flat YoY), in-line with the guide. Q2 was guided to ~$4.1B (+4% YoY) — the first positive YoY non-AI print guided in several quarters, suggesting modest stabilization. Status: Resolved in line.

OpenAI XPU program disclosure. OpenAI was confirmed as the sixth named XPU customer with explicit one-gigawatt 2027 deployment.

Resolved positively

What to watch into next quarter

Q2 FY2026 AI semiconductor revenue vs. the $10.7B (+140% YoY) guide. A beat with a Q3 step-up toward $13B+ would validate the 2027 $100B+ trajectory's required exit-rate; an in-line print with flattening sequential adds would suggest the curve is shallower than the headline implies.

Q2 consolidated gross margin landing vs. the 77% flat-QoQ guide. Management guided flat sequentially despite further XPU mix shift — whether margin holds at 77% or compresses tests the system-sales-margin-dilution thesis and sets the 2027 EBITDA model.

Seventh XPU customer or any additional named hyperscaler. A new customer would confirm the pipeline is genuinely structural and push the implied 2027 customer count past management's prior bounded LLM-frontier-developer framing.

Networking as % of AI revenue in Q2 actuals vs. the 40% guide. Whether networking sustains 40% (i.e. ~$4.3B in Q2) or settles back to the 33% range tests whether the Tomahawk 6 / 1.6T DSP architectural share gain is structural or pull-forward.

Infrastructure software re-acceleration vs. the +9% Q2 YoY guide. Q1 at +1.4% was a sharp deceleration; the +9% Q2 guide demands a step-up that the underlying VMware annuity has not yet shown. A miss here would invalidate the "low double-digit FY2026" framing.

Non-AI semiconductor follow-through on the +4% Q2 guide. The first guided positive YoY non-AI print in multiple quarters; a beat or sustained recovery would add a second growth lever beyond AI.

Any supply-chain re-disclosure. Management's "supply secured through 2028" is now the gating assumption behind the $100B+ 2027 figure. Any commentary on incremental wafer, HBM, or substrate constraints — or any indication that secured capacity has been expanded further — directly moves the multi-year revenue ceiling.

Sources

Broadcom Q1 FY2026 press release (8-K exhibit, filed 2026-03-04): https://www.sec.gov/Archives/edgar/data/1730168/000173016826000011/avgo-02012026x8kxex99.htm
Broadcom Q1 FY2026 earnings call prepared remarks and Q&A.

Get the next brief, free.

We publish analyst-grade earnings briefs the same day or morning after every call — headline numbers, segment KPIs, Q&A highlights, and tone analysis. Free during beta.

This is not investment advice.