OpenAI’s Head of Compute Infrastructure, Sachin Katti, says artificial intelligence will soon be designing the very chips and systems it needs to run.

The semiconductor supply chain that powers OpenAI’s ambitions is the same one that crypto miners, blockchain validators, and GPU-hungry DeFi protocols compete for. When the largest AI company on the planet starts building its own chips, the ripple effects hit every industry that depends on advanced silicon.

From Nvidia dependency to custom silicon

Katti, a former Intel executive, joined OpenAI in November 2025 to lead the company’s compute infrastructure efforts. His mandate is clear: reduce OpenAI’s reliance on off-the-shelf Nvidia GPUs while scaling compute capacity to meet what the company describes as demand in the tens of gigawatts.

The custom chip journey started in October 2025, when OpenAI partnered with semiconductor giant Broadcom to co-design specialized AI accelerators. That collaboration bore fruit in June 2026 with the unveiling of Jalapeño, OpenAI’s first custom chip built specifically for optimizing large language model inference workloads.

Jalapeño is designed to make inference cheaper and faster. Katti’s bigger claim is that AI itself is now assisting in the design of these chips, which could dramatically accelerate the iteration cycle for future hardware generations.

Why crypto markets should pay attention

OpenAI still heavily relies on Nvidia GPUs despite its custom silicon ambitions. It also maintains partnerships with AMD. But the strategic direction is unmistakable: vertical integration of the hardware stack, giving OpenAI better pricing leverage and less exposure to GPU supply crunches.

For crypto miners, particularly those running GPU-based operations for coins like Ethereum Classic or various AI-crypto hybrid networks, OpenAI’s chip strategy introduces a new variable. If custom AI chips reduce demand pressure on consumer and enterprise GPUs, miners could see improved availability and lower prices.

OpenAI’s massive power consumption, targeting tens of gigawatts, competes directly with mining operations for energy resources and data center capacity. In regions where both industries are expanding, that competition could drive up costs for everyone.

AI designing AI hardware: what it actually means

Katti’s assertion that AI is beginning to assist in chip design isn’t as futuristic as it sounds. Google has been using machine learning for chip layout optimization since 2021. What’s new is the scale of ambition: OpenAI isn’t talking about AI tweaking transistor placement, but AI participating in the fundamental architecture decisions of the systems it will eventually run on.

Investors watching the intersection of AI and crypto infrastructure should track two things. First, whether OpenAI’s custom chip production actually reduces pressure on the broader GPU market, which would benefit decentralized compute networks that rely on commodity hardware. Second, whether the energy demands of OpenAI’s expanding compute footprint create regulatory pressure that spills over into crypto mining policy, particularly in energy-constrained jurisdictions where both industries are competing for the same megawatts.