BusinessFriday, June 26, 2026· 2 min read

OpenAI, SpaceX and Others Push AI Forward With Custom Chips

TL;DR

A growing wave of companies is building custom AI chips to reduce reliance on a single supplier and make AI infrastructure more resilient. OpenAI’s planned Jalapeño inference chip, built with Broadcom, signals a broader shift toward more competition, efficiency, and flexibility in the AI hardware stack.

Key Takeaways

  • 1OpenAI is developing a custom inference chip called Jalapeño with Broadcom.
  • 2Major players including Google, Apple, and SpaceX are also investing in their own AI silicon.
  • 3Custom chips can reduce single-supplier risk and help companies optimize hardware for their specific AI workloads.
  • 4More competition in AI chips could improve availability, efficiency, and long-term innovation beyond Nvidia’s dominant ecosystem.

Nvidia has led the AI chip market for years, but a new wave of custom silicon efforts suggests the ecosystem is becoming more diverse and resilient. OpenAI’s planned Jalapeño inference chip, developed with Broadcom, is part of a broader move by major technology companies to tailor hardware to their own AI needs.

Why custom AI chips matter

Companies such as Google, Apple, SpaceX, and now OpenAI are looking to reduce dependence on a single supplier while improving performance for specific workloads. Inference chips are especially important because they power the real-world use of AI models after training, from chatbots to enterprise tools and consumer products.

  • Greater resilience: More chip options can reduce bottlenecks and supply risks.
  • Better optimization: Custom designs can be tuned for each company’s AI systems and products.
  • Stronger competition: A broader hardware market can encourage faster innovation and better pricing over time.

This shift does not diminish Nvidia’s achievements, but it does point to a healthier and more competitive AI infrastructure landscape. As more organizations invest in specialized chips, the AI industry could gain the capacity, efficiency, and flexibility needed to support the next generation of AI applications.

Get AI Wins in Your Inbox

The best positive AI stories delivered to your inbox. No spam, unsubscribe anytime.