Skip to main content

Broadcom Bets Big on AI Chips, Aims for $100B Revenue by 2027

Broadcom Doubles Down on Custom AI Chips

In a market long dominated by NVIDIA, Broadcom is emerging as a serious contender with its aggressive push into custom AI chips. During Wednesday's earnings call, CEO Hock Tan dropped a bombshell prediction: the company expects its AI chip business to surpass $100 billion in annual revenue by 2027.

That eye-popping forecast sent Broadcom shares soaring more than 4% in after-hours trading. It's not just Wall Street taking notice - tech giants are lining up for Broadcom's specialized silicon.

The Custom Chip Advantage

While NVIDIA rules with its versatile GPUs, Broadcom is carving out a different path. "We're not trying to be everything to everyone," Tan explained. "Our strength comes from building exactly what each customer needs."

This tailored approach appears to be paying off. Alphabet, Microsoft, Amazon and Meta are collectively planning to spend at least $63 billion on AI infrastructure by 2026 - much of it going toward custom solutions.

Major Deals in the Pipeline

Tan revealed several key partnerships that demonstrate Broadcom's momentum:

  • Anthropic: Expected to receive 1 gigawatt of Broadcom's TPUs by 2026, scaling to 3 gigawatts in 2027
  • OpenAI: Set to get its first batch of custom Broadcom chips in 2027, exceeding 1 gigawatt capacity
  • Meta: Contrary to rumors, Tan confirmed Meta's custom accelerator program is "full steam ahead" with shipments already underway

The CEO also announced a new $10 billion stock buyback program, signaling confidence in the company's financial future.

Key Points:

  • Broadcom projects $100B annual AI chip revenue within three years
  • Custom chip strategy winning business from tech titans
  • Major deals with Anthropic and OpenAI showcase growing influence
  • Meta partnership remains strong despite market speculation

Enjoyed this article?

Subscribe to our newsletter for the latest AI news, product reviews, and project recommendations delivered to your inbox weekly.

Weekly digestFree foreverUnsubscribe anytime

Related Articles

NVIDIA Bets Big on Groq Tech for Next-Gen AI Chips, Wins OpenAI Back
News

NVIDIA Bets Big on Groq Tech for Next-Gen AI Chips, Wins OpenAI Back

NVIDIA is shaking up the AI chip market with a powerful new partnership. The tech giant plans to unveil processors featuring Groq's lightning-fast language processing technology at next month's GTC conference. In a major coup, OpenAI has signed on as lead customer after briefly flirting with competitors. This move signals NVIDIA's determination to dominate the crucial AI inference market as computing demands evolve.

February 28, 2026
AI chipsNVIDIAGroq
News

Memory Chip Prices Hit Record Highs as AI Demand Overwhelms Supply

The first quarter of 2026 has seen memory chip prices nearly double, far exceeding industry forecasts. With AI servers consuming massive storage capacity, DRAM spot prices have skyrocketed by 369%, creating ripple effects across consumer electronics and stock markets. Major manufacturers are prioritizing high-profit AI components, leaving traditional devices scrambling for supply.

March 5, 2026
memory chipsAI hardwaretech industry
News

NVIDIA Defies Skeptics with Stellar $120 Billion Profit Year

Silencing doubts about an AI slowdown, NVIDIA just posted financial results that crushed expectations. The chipmaker's quarterly revenue hit $68.13 billion, powered by a 75% surge in its data center business. With annual profits reaching $120 billion, NVIDIA continues to dominate as tech's most valuable company while reshaping global AI infrastructure.

February 26, 2026
NVIDIAAI chipstech earnings
China's GLM-5 AI Model Breaks New Ground with Domestic Chip Support
News

China's GLM-5 AI Model Breaks New Ground with Domestic Chip Support

Zhipu Technology's GLM-5 AI model has made waves with its latest upgrades, now fully supporting seven major Chinese chip platforms. The model boasts a staggering 744 billion parameters and leads globally in programming agent capabilities. While user demand temporarily overwhelmed servers, the company has responded with compensation measures. Key innovations include a dynamic attention mechanism and new reinforcement learning algorithms that significantly boost performance.

February 23, 2026
AI innovationChinese techmachine learning
News

Memory Prices Set to Climb Through 2026 as AI Hunger Grows

SK Hynix warns that memory chip prices will keep rising through 2026, fueled by insatiable AI demand. While PC and mobile markets may cool, server needs remain strong. The company reports tight inventories across DRAM and NAND products, giving suppliers pricing power. Customers are learning the hard way that panic buying only worsens shortages.

February 22, 2026
semiconductorsmemory chipsAI hardware
News

Chinese Tech Firms Hit Milestone with AI Chip and Model Integration

In a significant step for China's tech independence, Moortu's MTT S5000 GPU has achieved full compatibility with Zhipu's advanced GLM-5 AI model. This breakthrough means domestic hardware can now fully support cutting-edge AI development without relying on foreign technology. The integration promises faster processing and lower costs for Chinese businesses adopting AI solutions.

February 12, 2026
AI chipsChinese technologyhardware innovation