Joining me on today's call are Hock Tan, President and CEO, Charlie Kawwas, President, Semiconductor Solutions Group, and Ram Velaga, President, Infrastructure Software Group. In our fiscal Q2 2026, total revenue reached a record $22.2 billion, up 48% year-on-year, above our guidance on strength in AI semiconductors. Q2 operating margin was a record 67%, and adjusted EBITDA was a record 69% of revenue, which was above our guidance. Even as our revenue scales up massively, driven by AI, our operating and EBITDA margins remain strong and stable.

Turning to semiconductors, Q2 revenue was a record $15 billion, as I said before, as we grew 79% year-on-year. Driving this growth was AI semiconductor revenue at a record $10.8 billion, up 143% year-on-year and above our outlook. In the second half of 2026, we expect AI semiconductor revenue to double from the first half we shipped this year. Consistent with this trend, in Q3, we expect AI semiconductor revenue to accelerate to $16 billion, up over 200% year-on-year.

For the full year 2026, we expect to achieve AI semiconductor revenue of $56 billion, up approximately 180% from fiscal 2025. We expect this momentum to continue into fiscal year 2027 and reiterate our AI semiconductor revenue guidance to be in excess of $100 billion. We expect AI semiconductor revenue growth to continue in fiscal 2028, based on the following initiatives with our six core customers. We have a contractual commitment to deploy 1.3 GW in 2027 as part of the larger 10 GW that by 2029 agreement we announced last year.

What went well
  • Record total revenue of $22.2 billion, up 48% year-on-year and above guidance on AI semiconductor strength.
  • Record operating margin of 67% and record adjusted EBITDA at 69% of revenue, above guidance, even as revenue scaled.
  • Record semiconductor revenue of $15 billion, up 79% year-on-year, driven by AI.
  • Record AI semiconductor revenue of $10.8 billion, up 143% year-on-year and above outlook; AI networking was almost 40% of AI revenue.
  • AI semiconductor bookings exceeded $30 billion in the quarter against $10.8 billion shipped, taking visibility into 2028.
  • Raised full-year FY2026 AI semiconductor guidance to $56 billion (up ~180% from FY2025); reiterated over $100 billion of AI semiconductor revenue in 2027 (~10 GW).
  • Secured multi-generation customer agreements: Google (multi-gen TPU), Anthropic (+5 GW from 2027), Meta (3 GW MTIA through 2028), OpenAI (1.3 GW in 2027 of the 10 GW by 2029).
  • Launched the AI XPU platform with Apollo, Blackstone and others to deploy over 20 GW through 2028; first tranche valued at $35 billion.
  • Non-AI semiconductors returned to growth (+6% YoY, bookings over $6 billion) signaling cyclical recovery; software up 9% with VCF 9.1 driving an accelerated Q3 (+31%).
  • Record free cash flow of $10.3 billion (46% of revenue); semiconductor operating margin 62%, up 460 basis points year-on-year.
What went wrong
  • Consolidated gross margin fell 230 basis points year-on-year to 77.1% as semiconductors became a larger mix; Q3 gross margin guided down to ~74% on AI mix.
  • Wireless declined seasonally within non-AI semiconductors.
  • Inventory days rose to 86 (from 68 in Q1) to secure supply for second-half AI growth.
  • CFO transition: Kirsten Spears retiring June 12 after 12 years, with Amie Thuener incoming.

Guidance Changes

MetricPeriodCurrent guidance
Consolidated revenueQ3 FY2026$29.4B, up 84% YoY
Semiconductor revenueQ3 FY2026~$20.5B, up 124% YoY
AI semiconductor revenueQ3 FY2026$16.0B, up over 200% YoY
Infrastructure software revenueQ3 FY2026~$8.9B, up 31% YoY
Non-AI semiconductor revenueQ3 FY2026~$4.5B, up 12% YoY
Consolidated gross marginQ3 FY2026~74% on higher AI revenue mix (not a structural semi-margin change)
Operating marginQ3 FY2026~67%, flat sequentially
Adjusted EBITDA marginQ3 FY2026~68% of revenue
AI semiconductor revenueFY2026$56 billion, up ~180%
AI semiconductor revenueFY2027Reiterated in excess of $100 billion (~10 GW)
Non-GAAP tax rateQ3 / FY2026~16%

Performance Breakdown

MetricYoYNote
Total revenue +48% AI semiconductor strength across XPUs and networking.
AI semiconductor revenue +143% Insatiable XPU and networking demand; bookings over $30 billion in the quarter.
Semiconductor revenue +79% AI-driven acceleration to a record $15 billion (68% of total revenue).
Infrastructure software revenue +9% Strong VMware bookings; ARR up 17%; VCF 9.1 momentum.
Non-AI semiconductor revenue +6% Broadband, server storage and enterprise networking up; path toward full cyclical recovery.
Operating income +52% Record $14.9 billion; operating margin 67.3%, up 200 bps YoY as opex stayed flat.
Semiconductor operating margin +460 bps Strong operating leverage; margin 62%.
Consolidated gross margin -230 bps Semiconductors became a larger proportion of product mix.
Free cash flow record $10.3B (46% of revenue) Strong cash generation on scaling AI revenue.

Earnings Call Themes & Trends

TopicPrevious mentionCurrent periodTrend
2027 AI revenue outlookLine of sight over $100 billionReiterated over $100 billion (~10 GW), on the same 2x trajectory as 2H 2026; 2028 expected substantially higher
Customer commitmentsVerbal gigawatt targetsContractual multi-generation agreements (Google, Anthropic +5 GW, Meta 3 GW, OpenAI 1.3 GW 2027) disclosed via 8-Ks
AI financing / XPU platformNot presentNew AI XPU platform with Apollo/Blackstone to deploy over 20 GW through 2028; first $35 billion tranche launching via Apollo
Rack vs chip debateSystem/rack-sale discussionDefinitively a chips-only business - 'no rack, only chips'
AI networking mix~One-third to 40% of AI revenueAlmost 40% this quarter (stars aligned); normalized level closer to 30%
CFO leadershipKirsten Spears CFOKirsten Spears retiring June 12; Amie Thuener incoming CFO
Next-gen networkingTomahawk 6 (100 Tbps) shippingTaping out 200 Tbps (Tomahawk 7-class) switch this quarter; leading on CPO, 1.6 Tb DSPs, Jericho 3/4

Q&A Summary

Reconcile the $56 billion FY2026 AI number with 2x half-over-half growth, and is the next-18-month AI backlog now ~$200 billion?
First-half AI was ~$19 billion; doubling gives ~$56 billion for the year, which ties out. 2027 will grow roughly 2x again, easily exceeding $100 billion; on track if not stronger. Not guiding a specific backlog figure but the trajectory is intact.
Given the Google long-term agreement 8-K, can you add color on share/upside within that customer?
A very strong agreement reflecting a substantial dollar commitment; Broadcom wants every design but accepts some source diversity as Google's AI consumption grows.
What is driving gross margin down harder within semiconductors - XPU vs networking, rack vs chip?
It is mix (semi vs software, and lower-margin ASIC/TPU/wireless vs rich-margin AI networking), not structural. Semi margins are stable/solid; business is chips only - 'only chips.'
Is TAM/content per gigawatt rising a lot as compute content grows (per a competitor's comments)?
Dollars per gigawatt are relatively stable because fewer, higher-power, higher-ASP chips are used; it is the number of gigawatts that accelerates. 2028 expected to be a substantial step up from 2027.
Can you get incremental wafer/HBM supply if customers want upside, and would you use other foundries?
Supply is secured for 2026-2027 with 2028-2029 in progress; getting supply is not just about spending money, but customers have been coming for incremental supply and Broadcom can largely meet it.
Are the 2027 gigawatt targets changed versus last quarter's ~10 GW?
Still ~10 GW planned for 2027, intact and back-half loaded, setting up substantially more gigawatts in 2028.
Is the Anthropic deal backstopped by Broadcom chips, and will future deals be financed this way?
Corrected the premise - Broadcom is providing its TPU chips/compute capacity to Anthropic, not backstopping; the new XPU platform (Apollo/Blackstone) funds chips for LLM labs that need capital.
Will networking stay near 40% of AI revenue as custom ramps accelerate?
40% reflects a rare alignment of strong non-XPU networking sales plus XPU growth; the expected normalized share is closer to 30%.
Is AI demand shifting toward consumer-facing labs (Anthropic, OpenAI) and will enterprise create a second wave?
Enterprise AI is early, but enterprises mostly buy tokens/APIs from the same few frontier labs (Anthropic, OpenAI, Gemini), so demand funnels back to the same compute-capacity customers Broadcom serves - a sustainable, steepening trajectory.
Why were AI bookings ($30 billion) so large versus shipments this quarter?
Customers order early to secure long lead times (wafers, HBM, power), giving visibility now to 2028 (vs 2027 a quarter ago); demand is limited by power and infrastructure, not Broadcom components.

More on Broadcom Inc.

Reported 2026-06-03 · figures from the Broadcom Inc. Q2 2026 earnings call.

See how VectorShift works for your firm

Request Demo