Broadcom Q2 FY2026 Earnings Preview: $22B Revenue, $73B AI Backlog, and the June 3 Print That Could Redefine AVGO

What You'll Learn

• Why the June 3 print is the most-watched AI infrastructure earnings of the week and what 47 Wall Street analysts expect from the top line
• How Broadcom's custom ASIC business broke out in Q1 FY2026 and whether the $10.7 billion Q2 AI guide is actually conservative
• The customer concentration story behind the $73 billion XPU backlog (Google, Meta, OpenAI, Anthropic, Apple, ByteDance)
• What the options market is pricing for AVGO's post-earnings move, why 94% of analysts are Buy-rated, and the realistic upside vs downside scenarios heading into the report

The Setup: Why AVGO's June 3 Print Could Reset the AI Infrastructure Narrative

The Broadcom Q2 FY2026 earnings report is the single most important AI infrastructure print of the week, and arguably of the second quarter. The Santa Clara chip conglomerate will release results after the close on Wednesday, June 3, 2026, with the conference call scheduled for 5 p.m. Eastern. Wall Street is not just watching a number — it is watching whether Hock Tan can prove that the $73 billion XPU backlog and 106% AI revenue growth from Q1 FY2026 are structural rather than cyclical, and whether the $100 billion-plus 2027 AI revenue target the CEO floated in March is realistic or aspirational.

Broadcom enters the report as one of the Magnificent Seven's quietest power players. The stock is up roughly 115% from its 52-week low, hit an all-time high of $448.58 on May 29, 2026, and trades at a forward P/E that screens expensive on conventional metrics but cheap on AI-revenue acceleration. Wells Fargo, Morgan Stanley, UBS, and Susquehanna have all raised price targets in the last three weeks, with the Street's average target now around $483. The setup is the kind of asymmetric risk-reward that institutions hate to fade and that retail investors should approach with discipline.

The macro backdrop is unusually supportive. The S&P 500 just closed its eighth consecutive weekly gain, the ISM Manufacturing PMI hit 54.0% in May (a four-year high), and the Federal Reserve is widely expected to hold rates steady at the June 16–17 FOMC meeting, with the first real rate-cut window now pushed to the September meeting. That combination — strong macro, benign rates, and an AI capex super-cycle — is the exact environment where Broadcom tends to print upside. But it is also the environment where expectations have never been higher.

The AI Revenue Engine: From $8.4B in Q1 to a $100B+ 2027 Target

Broadcom's custom AI silicon business is no longer a side story. In Q1 FY2026 (reported March 4, 2026), the segment generated $8.4 billion in revenue, up 106% year-over-year, and represented the single largest growth driver in the company's $19.3 billion top line (up 29% YoY overall). Adjusted EPS landed at $2.05, beating the $1.88 consensus. More importantly, management raised the Q2 revenue forecast from $20.4 billion to roughly $22 billion and guided Q2 AI revenue to approximately $10.7 billion — a 27% sequential jump and a year-over-year increase that some analysts have put as high as 140%.

That $10.7 billion number is the key print to watch on Wednesday. If Broadcom can deliver on its own AI guide and then add a third-quarter commentary that points to $12 billion or more, it would represent a 60% annualized growth rate and validate the much more aggressive longer-term target Hock Tan laid out at the Q1 call. Tan told analysts in March that Broadcom's AI revenue trajectory puts the company on a path to exceed $100 billion in AI chip revenue by fiscal 2027 — a number that, if achieved, would represent roughly 40% of the entire custom AI ASIC market that sell-side analysts size at $60–90 billion in TAM by 2027.

The math behind that $100 billion target is straightforward but aggressive. Broadcom is targeting 20–25% market share of the custom AI accelerator space, with an XPU backlog of $73 billion spread across six named customers and shipping over an 18-month window. If even half of those orders convert at the gross margin profile that the segment has historically enjoyed (analyst estimates put AI silicon gross margins above 75%), the AI segment alone could be generating $40+ billion in gross profit by 2027. That is the kind of forward earnings power that justifies a forward P/E in the high-30s even at today's price.

The Customer Roster: Six Hyperscalers, $73 Billion, and the Concentration Story

The most under-appreciated part of the Broadcom bull case is the customer concentration, and it cuts both ways. The $73 billion XPU backlog is concentrated across six hyperscalers: Google, Meta, Apple, ByteDance, OpenAI, and Anthropic. That list is essentially the entire frontier-AI capex base of the Western and Chinese tech industries, and Broadcom has a piece of nearly all of it.

Google remains the anchor. Broadcom has co-designed every generation of Google's Tensor Processing Unit (TPU) since the original TPUv1 in 2015, and the partnership deepened materially in April 2026 when Anthropic signed a new agreement with Google and Broadcom for multiple gigawatts of next-generation TPU capacity. That deal was disclosed jointly by Anthropic and Broadcom in an April 6, 2026 official announcement, and effectively bakes in a multi-year revenue stream tied to Anthropic's Claude training and inference workloads. The Anthropic commitment, which was followed by Anthropic's $965 billion IPO filing in May, is one of the more important endorsements of the TPU-Broadcom design path that the market has seen.

Meta is the second pillar. On April 15, 2026, Broadcom announced a multi-year partnership with Meta Platforms for Meta's custom MTIA XPU chips, with one gigawatt of capacity as the initial commitment. Morningstar's analyst team called the deal a "bullish XPU confirmation" in a note published the same day. Apple's silicon relationship with Broadcom — covering modem, wireless, and now on-device AI accelerators — has been ongoing for years but took on new significance with the announcement of Apple's WWDC 2026 push into on-device generative AI. ByteDance, the TikTok parent, remains the largest non-US customer and is a critical test of the Chinese hyperscaler cohort, where the export-control environment adds an additional layer of complexity that does not affect US-based peers.

OpenAI joined the customer list in March 2026, becoming Broadcom's newest design partner (per Yahoo Finance's coverage of the deal) and providing the clearest signal yet that the ChatGPT operator is pursuing a multi-vendor silicon strategy rather than relying solely on Nvidia. The combination of Nvidia's $81.6 billion Q1 FY2027 print and the simultaneous expansion of Broadcom's customer base suggests the AI accelerator market is bifurcating into a "general-purpose GPU" tier dominated by Nvidia and a "custom ASIC" tier dominated by Broadcom, with hyperscalers hedging by maintaining both. The HBM memory chip supercycle sits inside both tiers — and Broadcom's networking silicon is one of the key bottlenecks for HBM-stacked GPU clusters, providing a second revenue stream tied to the same AI capex cycle.

Wall Street's Math: $22B Revenue, $2.40 EPS, and the 94% Buy-Rated Street

The consensus numbers heading into Wednesday are unusually well-defined. Of 37 analysts tracked by LSEG and FactSet who publish explicit revenue estimates, the median is $20.6 billion — but that number is already stale. Management's own guidance of $22 billion, issued at the Q1 FY2026 call on March 4, has effectively set the floor. A print at or above $22 billion would represent 47% year-over-year growth, the highest top-line growth rate Broadcom has posted in any quarter since the VMware acquisition closed in November 2023.

On EPS, the consensus is $2.40, with a range of $2.36 to $2.54 across the 37 publishing analysts. Adjusted gross margin is expected to come in at 76.0%, broadly in line with Q1's 76.5% and consistent with the mix shift toward higher-margin AI silicon. Adjusted operating margin should land in the 47–48% range, and the adjusted EBITDA margin of roughly 64% will be the number that institutional investors focus on as a proxy for the cash-generative power of the AI segment.

Sentiment on Wall Street is essentially one-sided. Of 47 analysts covering the stock, 44 rate it a Buy or Strong Buy, three rate it Hold, and zero rate it Sell. The average 12-month price target sits at $483.95, with a range of $360 (Zacks low end) to $630 (Zacks high end). The most recent target raises include Wells Fargo moving to $545 from $430, Morgan Stanley to $485 from $470, UBS, and Susquehanna. The Wells Fargo move is the most aggressive — it implies 22% upside from the May 29 all-time high of $448.58 and effectively prices in a Q2 beat-and-raise scenario plus an above-consensus Q3 guide.

The risk to that consensus is twofold. First, the stock has already absorbed much of the bull case — at $448 it trades at roughly 38x forward earnings, a multiple that compresses quickly if the AI revenue growth rate decelerates even by a few hundred basis points. Second, the LSEG consensus revenue number of $20.6 billion is materially below management's $22 billion guide, which means a meaningful subset of sell-side models may need to be revised higher in the days after the print. That revision cycle is what tends to drive the post-earnings drift, and it is the primary mechanism by which the stock can extend even after a clean beat.

Stock Setup, Options Pricing, and the 7.5% Implied Move

The options market is pricing Broadcom for a roughly 7.5% move on Wednesday, which translates to a trading range of approximately $415 to $482 from the May 29 close of $448.58. That implied move is wider than the stock's trailing four-quarter average post-earnings move of 5.2% but narrower than the 9.8% it printed after the Q1 FY2026 release in March, which was a clean beat that included the raised Q2 guide. A move of 7.5% would be a 32-point swing, which is consistent with the dollar-implied volatility priced into the straddle.

For context, the broader macro backdrop has been unusually supportive, with the S&P 500's eight-week winning streak and the ISM Manufacturing PMI's four-year high both creating a tailwind for cyclical AI-exposed names. The Federal Reserve's June 16–17 FOMC meeting, where the dot plot is expected to push the first rate cut to September, should not materially impact Broadcom's print but will shape the narrative around forward valuation. The May jobs report preview we published last week flagged the wage-growth component as the key variable for the Fed's path, and that data point is also relevant for the AI capex cycle because labor costs in semiconductor manufacturing are a key input.

The most asymmetric setup is a beat-and-raise with above-consensus Q3 guide. If Broadcom prints $22.5 billion in revenue, $2.45 in EPS, and guides Q3 to $24 billion-plus, the stock would likely gap to $500 on the open and could trade to $520–540 over the following week as sell-side models revise. Conversely, a clean in-line print at $22 billion with a cautious Q3 guide would likely trigger a 5–8% pullback to the $415–425 range, where the 50-day moving average sits. The downside scenario — a miss and a guide-down — is essentially priced out by the current consensus, which is one reason the options market is asymmetrically favoring upside calls.

VMware, Margins, and the Risks Wall Street Won't Ignore

The VMware integration is the variable that does not show up in the AI narrative but materially affects the operating story. Q2 FY2026 will be the second full quarter with VMware consolidated under Broadcom's software segment, and investors will be looking for evidence that the $69 billion acquisition is delivering on the cost synergies and revenue retention that Hock Tan promised at the deal close. The first-quarter integration showed stable VMware revenue but elevated churn risk among enterprise customers who viewed the acquisition as a price-hike trigger. The Q2 print will reveal whether that churn is intensifying, stabilizing, or reversing.

Margin guidance is the second underappreciated risk. AI silicon gross margins are above 75%, but the software segment is structurally lower, and the mix shift toward software (which was roughly 40% of revenue in Q1) is a slight gross margin headwind. If VMware's contribution to the consolidated business expands faster than the AI segment in Q2, the gross margin could come in 50–100 basis points below the 76.5% Q1 print. That is not a thesis-breaker, but it is a number that growth-focused investors tend to react to disproportionately.

Customer concentration is the third risk. Six hyperscalers represent essentially all of the $73 billion XPU backlog, and the loss of any one of them would be a material event. Google's TPU business is the most mature, Meta's MTIA is the newest, and OpenAI's commitment is the most recent. The company has not disclosed revenue concentration by customer, but most sell-side estimates put Google at 35–40% of AI segment revenue, Meta at 15–20%, and the remaining four customers splitting the rest. Any disclosure on Wednesday that materially shifts those concentration ratios — either up (more Google) or down (more diversification) — would be a meaningful print detail.

Tariff Tail Risk, Supply Chain, and the Macro Overhang

The one variable that has almost no sell-side model coverage but is real is tariff exposure. Broadcom manufactures a meaningful portion of its networking and AI silicon through TSMC in Taiwan, and the second Trump administration's tariff regime has introduced both direct chip tariffs and indirect supply-chain costs that are not fully reflected in the Q1 guidance. Hock Tan's commentary on tariffs will be one of the most-watched sections of the call, especially given that several competitors (Marvell, Nvidia) have already quantified tariff exposure in their own recent guidance. Any quantified tariff hit of 2–5% of revenue would be a meaningful negative for the multiple.

Supply-chain bottlenecks in HBM memory, advanced packaging, and CoWoS capacity are the second macro overhang. Broadcom's networking silicon pairs with HBM memory and CoWoS-packaged accelerators, and the same capacity constraints that have affected Nvidia and AMD throughout 2025–2026 also apply to Broadcom's AI segment. If Q2 commentary flags any softening in customer orders because of memory or packaging shortages, it would be a leading indicator that the AI capex cycle is hitting a near-term air pocket, even if the longer-term trajectory remains intact. The recent Apple WWDC 2026 push into on-device AI added another layer of HBM and packaging demand that is competing for the same constrained capacity.

The macro picture is supportive but not uncomplicated. The May 2026 ISM Manufacturing PMI at 54.0% is a clear positive for cyclical demand, and the FOMC's expected hold at the June 16–17 meeting keeps the cost of capital stable. But the labor market is showing signs of re-acceleration, which could keep wage-driven inflation sticky and push the Fed's first rate cut out of September. For Broadcom specifically, the May jobs report wage-growth number is the most relevant near-term macro input because it directly affects the Fed's reaction function.

The Bottom Line: How to Position Into Wednesday

Broadcom's Q2 FY2026 print is the most-watched AI infrastructure earnings of the week, and the setup is one of the more asymmetric in the Magnificent Seven. The $22 billion revenue guide, the $10.7 billion AI segment number, and the 106% Q1 AI growth rate are all in the consensus. The questions that will drive the stock are (1) does Q3 guidance come in above the implied $23–24 billion range, (2) does Hock Tan reaffirm or raise the $100 billion-plus 2027 AI revenue target, and (3) what is the customer concentration disclosure. A clean beat-and-raise with an above-consensus Q3 guide would likely push the stock to $500 and set up a run toward the $545 Wells Fargo target. A cautious print with an in-line Q3 guide would likely trigger a 5–8% pullback to the $415–425 range, where the technical support and 50-day moving average converge.

The asymmetric risk-reward favors a barbell: long-dated calls (October or January expiry) sized small enough to absorb a 5% post-earnings drawdown, paired with a defined-risk collar if the position size is material. The 7.5% options-implied move is reasonable but not generous given the size of the expected revision cycle, and the IV crush after the print is likely to be 30–40% for the front-month options. For investors with a 12-month horizon, any pullback to the $420–430 range is a high-quality add into a structural AI infrastructure compounder with $73 billion of locked-in backlog, 47% top-line growth, and 94% sell-side Buy ratings.

The trade is not without risk. The multiple is rich, the customer concentration is real, the tariff overhang is unresolved, and any deceleration in the AI capex cycle would compress the multiple quickly. But the combination of a structural $100 billion AI revenue target, a hyperscaler customer roster that is essentially the entire frontier-AI ecosystem, and a macro backdrop of stable rates and accelerating manufacturing PMI creates a setup where the bull case is well-defined and the bear case requires multiple things to go wrong simultaneously. For Broadcom, Wednesday is not just an earnings print — it is the print that will determine whether the AI infrastructure trade extends into the second half of 2026 or takes a breather.

Last Updated: June 02, 2026 | Source: Broadcom Investor Relations, Wall Street research, and market data compiled from public sources (Bloomberg, Reuters, CNBC, MarketWatch, Yahoo Finance, and analyst notes from Wells Fargo, Morgan Stanley, UBS, and Susquehanna).