The question surrounding Broadcom (NASDAQ:AVGO) heading into its next earnings report is not whether the company belongs among the AI trade’s biggest winners. That debate ended a long time ago. The more pressing question is whether a business already generating tens of billions in AI-related revenue can keep growing fast enough to justify a valuation that has briefly crossed $2 trillion.

Citi’s answer, for now, is yes. Analyst Atif Malik raised his price target on the stock from $475 to $500 on May 12, kept his Buy rating in place, and named Broadcom his top semiconductor pick for 2026. The timing was deliberate. Broadcom reports its fiscal second-quarter results on June 3, and Malik is going on record ahead of the print rather than reacting to it afterward, a meaningful signal of conviction given how far the stock has already traveled.

The Numbers Behind the Confidence

The bullish case starts with what Broadcom has already delivered. In its most recent quarter, the company posted revenue of $19.3 billion, up 30% year over year, with AI semiconductor revenue coming in at $8.4 billion, more than doubling from the same period a year earlier. Those are not the numbers of a company riding a temporary wave. They reflect a business that has become structurally embedded in how the world’s largest technology firms build their AI systems.

Management has guided for revenue of roughly $22 billion in the June quarter, representing nearly 47% year-over-year growth, with AI semiconductor revenue projected at $10.7 billion. Analysts are modeling earnings per share of around $2.32, compared with $1.60 in the same quarter last year.

Underpinning that confidence is something rare in the semiconductor industry: visibility. Broadcom is currently sitting on a $73 billion backlog in AI-related orders spanning custom accelerators, networking chips, and related components, with deliveries expected over the next 18 months. Much of its near-term revenue is effectively already contracted, which insulates it from the demand volatility that has historically made chip stocks difficult to own through earnings seasons.

What Broadcom Actually Does and Why It Matters

Broadcom’s position in AI is fundamentally different from Nvidia’s, and that distinction is worth understanding.

Rather than selling general-purpose chips into a broad market, Broadcom works directly with a concentrated group of hyperscale technology companies to design custom silicon tailored to specific workloads. Its key relationships include Google, Meta, Anthropic, and OpenAI. Google’s tensor processing units, which power much of the AI infrastructure behind its core products, are developed in close partnership with Broadcom. The company also supplies the networking chips that connect thousands of processors inside large AI clusters, enabling them to function as a coherent system.

This is a harder business to displace than it might look. Custom chip design requires years of collaboration, deep integration into a customer’s software architecture, and institutional knowledge that cannot be transferred quickly. Citi specifically cited Broadcom’s five-year agreement with Google as protection against the risk of customers eventually going it alone, arguing it would be extremely difficult for any alternative supplier to catch up technologically.

Beyond semiconductors, the VMware-anchored software division continues generating substantial recurring revenue from large enterprises, providing earnings stability that pure-play chip companies cannot match.

A Long Runway, With Lofty Expectations Already Baked In

Malik’s long-term projections are ambitious even by the standards of the current AI cycle. He estimates that AI revenue, currently representing roughly half of Broadcom’s total sales, will grow to account for approximately 81% of the business by late fiscal 2028. Total AI sales are projected at $115 billion in 2027, rising to $180 billion in 2028, figures that would make Broadcom’s AI division alone larger than most chip companies’ entire revenue base today. CEO Hock Tan has himself targeted more than $100 billion in AI sales by 2027, so the ambition is not simply an analyst overlay on conservative guidance.

The caveat is that the stock reflects much of this already. Trading above 80 times trailing earnings near a $2 trillion market cap, there is limited tolerance for execution stumbles. A recent report that OpenAI is facing an $18 billion financing gap tied to its custom chip partnership with Broadcom introduced fresh uncertainty, a reminder that even the strongest infrastructure thesis carries real funding and delivery risk at this scale.

Why June 3 Matters Beyond Broadcom Alone

Broadcom’s earnings have become something of a proxy for the health of hyperscaler AI spending broadly. Strong results tend to lift confidence across the semiconductor sector. Cautious management commentary tends to do the opposite.

With AI infrastructure valuations still pricing in years of aggressive capital expenditure from the world’s largest technology companies, the June 3 report will be read not just as a Broadcom story, but as a referendum on whether that spending trajectory is holding. Citi’s decision to raise its target ahead of the print, rather than after, suggests the firm believes the answer will be yes.

Benzinga Disclaimer: This article is from an unpaid external contributor. It does not represent Benzinga’s reporting and has not been edited for content or accuracy.