
Broadcom's stock has declined nearly 20% from its June peak but the company forecasts AI semiconductor revenue will exceed $100 billion in 2027, up sharply from $10.8 billion in Q2 2026.
This growth is driven by adoption of Broadcom's custom AI chips—application-specific designs that are more cost-effective than general-purpose GPUs for AI hyperscalers.
At a forward earnings multiple of 20 times for 2027, the market may not yet be fully pricing in this expansion.
What happened
Broadcom's stock has fallen nearly 20% from its June peak despite rising about 14% year to date in 2026. The company expects its AI semiconductor revenue to exceed $100 billion in 2027, up from $10.8 billion in the second quarter (or about $43 billion annualized).
Why it matters
Broadcom's custom AI chips—application-specific integrated chips designed for particular workloads—are emerging as cost-effective alternatives to general-purpose GPUs for AI hyperscalers. The company is expanding beyond its current handful of clients, with major new orders expected to hit in 2027. At a forward earnings multiple of 20 times for 2027, investors may not yet be pricing in the full extent of this growth.
What to watch
Broadcom's custom chip orders are expected to accelerate in 2027, starting to roll in and driving revenue and profit growth. The company has already helped Alphabet develop the Tensor Processing Unit (TPU), which is becoming increasingly popular and driving Google Cloud's growth.
Ask the AI about this article →
Broadcom's upcoming AI revenue surge hinges on a strategic shift in how AI hyperscalers procure computing infrastructure. Rather than relying solely on general-purpose GPUs—which offer flexibility but waste capacity on narrowly defined, predictable workloads—major cloud providers are partnering with Broadcom to design application-specific integrated chips tailored to their exact needs. This approach, already validated by Alphabet's success with the Tensor Processing Unit, delivers both cost savings and performance advantages for hyperscalers managing enormous capital budgets. The company's current AI semiconductor run rate of roughly $43 billion annualized will need to more than double to reach the $100 billion forecast for 2027; the article indicates this growth is driven by orders from several new hyperscaler clients beginning to roll out in 2027.
From a valuation perspective, the market appears to be underpricing this transition. While Broadcom trades at 34 times current-year earnings—a premium multiple—the forward multiple for 2027 compresses to 20 times, suggesting that only a portion of the expected growth has been capitalized into the stock price. The fact that the stock has fallen nearly 20% from its June peak despite Broadcom's unchanged growth guidance creates a potential disconnect: investors may be reacting to short-term sentiment or broader market weakness rather than to fundamental deterioration in the company's AI opportunity.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
U.S. markets ended August higher, with the S&P 500 up 2.6% and the Nasdaq up 3.9%

Palantir Technologies stock has posted multi-year gains, including an 11x return over 3 years

CrowdStrike shares rose 5.77% on Monday, driven by investor excitement over its AI-powered cybersecurity growt…

Nvidia paused some deals under its AI financing program, which offered credit support to smaller AI cloud comp…

NVIDIA is now acting as a financier for its own demand, helping raise over $500 billion of third-party capital…

GE Vernova unveiled its medium-voltage uninterruptible power supply (MV-UPS) on August 24
