Broadcom Sees $100B Artificial Intelligence Chip Run Rate By 2027

Broadcom Lifts Guidance As AI Demand Rewrites The Growth Story

Broadcom’s latest outlook reset investor expectations for how quickly artificial intelligence infrastructure spending is turning into chip and networking revenue. Chief executive Hock Tan said the company has clear visibility into a much larger opportunity, as customers expand from building training clusters to deploying products, agents, and enterprise services at scale. The comments landed alongside a quarterly beat, a stronger second-quarter forecast, and a fresh $10 billion buyback authorization, a mix that tends to move semiconductor valuations fast when growth is perceived as durable.

The company posted fiscal first-quarter revenue of $19.31 billion, up 29% from a year earlier, and adjusted earnings of $2.05 per share. Broadcom guided to about $22.0 billion in second-quarter revenue, well above prior expectations, with artificial intelligence semiconductor revenue projected at roughly $10.7 billion for the quarter. In the first quarter, AI-related revenue more than doubled to $8.4 billion, signaling that the current cycle is being driven by both compute and the plumbing that connects it.

A $100B Ambition Anchored In Custom Silicon And Connectivity

Tan told investors the company is targeting an annualized scale that would place Broadcom among the largest beneficiaries of the data center buildout. The figure he referenced was $100B, a level that implies a step-change from current quarterly performance and would require continued expansion in custom accelerators and high-speed networking. Broadcom’s positioning is less about selling general-purpose GPUs and more about co-designing chips with large customers, then pairing those designs with switching, interconnect, and related components.

That mix matters for revenue quality. Custom projects can be sticky once a platform is in production, and networking content rises as clusters grow and bandwidth requirements increase. Broadcom has highlighted that networking is becoming a larger slice of its AI business, with an internal expectation that it can represent a sizable share of AI revenue as deployments broaden.

Customer Concentration, Now With A Sixth Name Coming Into View

The company’s AI customer list has expanded, and management has described a path to six hyperscaler-scale buyers. Several of the largest cloud and platform groups are simultaneously ramping spending on compute, memory, power, and networking, and Broadcom’s strategy is to capture wallet share where customers want differentiated silicon and predictable supply.

Tan also pointed to additional large-scale demand from a fast-growing AI model developer, describing compute requirements in gigawatt terms. Power-based framing is a signal that the discussion has moved beyond unit shipments to full system deployments, where each incremental model release can translate into incremental racks, switches, and optical connectivity.

Supply Chain Control Is The Quiet Constraint Behind The Narrative

A central question for markets is whether the bottleneck is demand or deliverability. Broadcom’s confidence rests partly on its ability to secure capacity, packaging, and the supply chain steps that often determine how much can ship in any given quarter. When management says it “sees” line-of-sight, investors typically interpret that as a combination of committed customer roadmaps and production readiness across the ecosystem.

If supply holds, Broadcom can convert the current order environment into a smoother revenue ramp. If supply tightens, the near-term impact would likely show up as deferred shipments rather than lost demand, but the market often penalizes timing risk because it can distort quarterly comparisons and compress multiples.

How This Hits Markets Beyond One Stock

The immediate channel is equities. A stronger semiconductor guidance print can lift the broader complex, especially names tied to data centers and networking. It can also tighten the performance gap between general compute plays and infrastructure plays, because Broadcom’s outlook argues that the connectivity layer remains a growth engine, not a second-order beneficiary.

Rates and credit matter through a different mechanism. If AI-led capex remains strong, it can support growth-sensitive assets, but it can also keep upward pressure on longer-end yields if investors expect sticky investment and energy demand. In credit, stronger cash generation plus buybacks can be supportive for issuer sentiment, though aggressive capital returns can raise questions if the cycle turns and spending requirements increase.

Base Case, Upside Scenario, Downside Scenario

The base case is that Broadcom continues to execute on second-quarter guidance, keeps AI semiconductor revenue rising, and sustains momentum in data center networking. Under this path, the stock’s support comes from repeatable demand and the ability to meet customer timelines, while the buyback provides an additional floor during volatile tape.

The upside scenario requires two triggers: additional custom silicon wins moving from design into volume production, and networking attach rates rising faster than expected as deployments shift from training-heavy to inference-heavy fleets. If those triggers hit, the company can pull forward growth, and the market may re-rate the name as a multi-year compounder rather than a single-cycle beneficiary.

The downside scenario is that customer build schedules slip, or that competitive pressure compresses pricing on custom accelerators and networking components. A second downside trigger is any supply-chain disruption that delays deliveries into key quarters, which can turn a strong demand story into uneven reported results and higher volatility around earnings.

What Investors Will Track Into 2027

The 2027 timeline implied by management’s ambition will be judged on measurable milestones: customer count at scale, quarterly AI revenue progression, and evidence that networking continues to rise as a share of the artificial intelligence stack. Investors will also watch how software and non-AI segments behave, because diversification can cushion results if one end market softens.

For now, Broadcom’s message is that the demand curve is broadening, not narrowing, and that the company is positioning its chip and connectivity portfolio to capture a larger share of the infrastructure that powers the next wave of deployed AI applications.