Broadcom's AI Chip Surge: What Sales Teams Should Do Right Now
Broadcom blew past expectations on heavy AI chip demand and new partnerships with major tech players. The company posted adjusted EPS of $1.95 versus $1.42 a year ago and booked $18.02 billion in revenue, up 28% year over year.
The standout: AI semiconductor sales jumped 74%. Semiconductor solutions rose 35% and infrastructure software gained 19%. Broadcom guided next quarter revenue to $19.1 billion and expects AI semiconductor revenue to reach $8.2 billion as custom accelerators and Ethernet AI switches gain traction.
Deals with Alphabet and OpenAI deepen Broadcom's role in core AI workloads, spreading risk and supporting durable demand. For sellers, this is a clear signal: budgets are moving to AI infrastructure now.
The numbers that matter
- Adjusted EPS: $1.95 vs. $1.42 last year
- Revenue: $18.02B (+28% YoY), above estimates
- AI semiconductor sales: +74% YoY
- Semiconductor solutions: +35%; Infrastructure software: +19%
- Next quarter guide: $19.1B revenue
- Projected next quarter AI semiconductor revenue: $8.2B (custom accelerators, Ethernet AI switches)
For reference and context, see Broadcom's investor updates and docs on large-scale AI chips: Broadcom Investor Relations and Google Cloud TPU overview.
Why this matters for sales
Budgets are shifting to AI infrastructure that reduces training time, lowers inference cost, and increases throughput. That shift opens doors across cloud, data platforms, networking, and security.
Use this momentum to reframe your product around AI outcomes: faster models to production, lower cost per query, higher uptime under load. Translate features into measurable time, cost, and risk wins tied to AI roadmaps.
Plays to run this quarter
- Refresh your ICP: prioritize accounts building or scaling AI platforms (cloud, data-center operators, top-tier enterprises with LLM roadmaps).
- Multi-thread deals: involve infra, data, ML platform, and finance early. AI spend is cross-functional.
- Lead with "time-to-value": show how you reduce training cycles, latency, or infra overhead in 30-90 days.
- Use social proof: reference momentum with Alphabet and OpenAI to lower perceived risk and speed procurement.
- Bundle adjacencies: pair your offer with networking, observability, or security to capture budget while it's being allocated.
- Create urgency: capacity is tight. Position pilots and POCs now to secure resources before budgets lock.
Prospect angles worth testing
- Cloud providers and hyperscalers expanding AI services
- Enterprises building internal AI platforms (training and high-volume inference)
- Data-center builders, OEMs, and integrators modernizing networking for AI workloads
- Teams evaluating custom accelerators vs. general-purpose GPUs
Qualification questions that surface budget
- Which AI workloads are you scaling in the next two quarters-training, inference, or both?
- Where are you constrained today-throughput, bandwidth, or cost per query?
- Are you evaluating custom accelerators or Ethernet-based AI networking to hit your targets?
- What's the cost of waiting one quarter (SLA risk, latency targets, unit economics)?
- Who owns the AI infra budget, and what milestones release funds?
Objections you'll hear-and simple responses
- "Budgets are frozen." → Tie your value to AI OKRs already funded; quantify cost-per-inference or time-to-train savings.
- "We're standardizing on one vendor." → Many leaders are spreading risk across partners; propose a low-risk pilot tied to a single workload.
- "Let's revisit next year." → Capacity and lead times are tight; lock timelines now to avoid slipping AI launch dates.
What to watch next
- Broadcom's performance vs. the $19.1B revenue guide
- Adoption of custom accelerators and Ethernet AI switches as a share of AI spend
- Signals from other chip and infra vendors doubling down on AI-use these in your outreach for timely, relevant talk tracks
If you want quick, practical upskilling to sharpen your AI talk track and objection handling, explore AI courses by job.
Your membership also unlocks:
