Broadcom CEO Sees AI Chip Sales Crossing $100 Bn by 2027

Broadcom Inc. has significantly upgraded its long-term outlook, with Chief Executive Officer Hock Tan announcing that the company now expects its artificial intelligence (AI) chip revenue to exceed $100 billion in fiscal 2027, reflecting accelerating demand for custom silicon in hyperscale data centres.

Key Highlights

• Broadcom Inc. expects AI chip revenue to exceed $100 billion by 2027.
• CEO Hock Tan cites strong hyperscaler demand and improved long-term supply chain visibility.
• AI semiconductor revenue surged 106% year-on-year, reaching $8.4 billion in latest quarter.

Chief Executive Officer Hock Tan emphasised that Broadcom’s visibility into 2027 results has sharply improved, expressing confidence in the company’s long-term growth path for AI silicon:

“Our visibility in 2027 has dramatically improved. Today, in fact, we have line of sight to achieve AI revenue from chips in excess of $100 billion in 2027,” Hock Tan said on a post-earnings call with analysts.

He added that Broadcom has “also secured the supply chain required to achieve this,” underscoring the company’s preparations to support multi-year capacity needs for its largest customers. 

Strong Results and Revenue Outlook

Broadcom reported robust financial results for the quarter ended Feb. 1, 2026:

• Total revenue: $19.31 billion, up 29 % year-on-year and above Street estimates.
• AI-related semiconductor revenue: $8.4 billion, more than double the prior year.
• Second-quarter guidance: anticipating approximately $22 billion in revenue, well above analyst consensus of around $20.56 billion.

Tan said AI demand remains a principal growth engine, with Broadcom’s custom AI accelerators and networking silicon driving much of the upside. The company also unveiled a $10 billion share repurchase programme through 2026, a sign of confidence in its cash generation and long-term strategy.

Also Read:South Korea's KOSPI Crashes 13% Amid Hormuz War Fears

Custom AI Chips and Hyperscaler Partnerships

Unlike rivals that build whole GPUs, Broadcom specialises in application-specific integrated circuits (ASICs) and bespoke AI processing units (XPUs) for major tech customers.

• Broadcom engineers are helping customers like Google develop Tensor Processing Units (TPUs) and collaborating with OpenAI on custom processors that are slated for volume deployment in 2027.
• The company expects to deliver 1 gigawatt of TPU compute capacity to Anthropic in 2026, rising to over 3 gigawatts in 2027 — a metric used by the industry to gauge scale in AI compute.

Analysts such as Gil Luria of D.A. Davidson noted that issuing guidance out to 2027 — beyond the usual near-term outlook — indicates exceptionally strong demand visibility for Broadcom’s silicon and networking solutions.

Stock Market Reaction & Competitive Position

Broadcom’s share price reacted positively to the forecast, climbing more than 4 % in extended trading as investors digested the stronger guidance and buyback announcement.

The projection places Broadcom in a more competitive position within the AI silicon landscape traditionally dominated by companies like Nvidia, which has reported multi-gigawatt AI deal volumes itself. Yet Broadcom’s focused approach on custom silicon and networking — including its patented stacked design technologies — is seen as a differentiator in hyperscale deployments.