Wafer, an AI startup, is training models to automatically optimize code for custom chip architectures using reinforcement learning on augmented Claude and GPT models. The company is working with Amazon and AMD to help them compete with Nvidia's software-driven chip dominance. By democratizing code optimization, Wafer could accelerate adoption of custom silicon across the industry and erode Nvidia's $4 trillion monopoly.
Strategy
AI Could Democratize One of Tech's Most Valuable Resources
AI startup Wafer is using reinforcement learning to automate code optimization for custom chips, potentially undermining Nvidia's $4 trillion dominance by making custom silicon architectures competitive for Amazon, AMD, and others.
Wednesday, April 15, 2026 12:00 PM UTC2 MIN READSOURCE: WIRED AIBY sys://pipeline
Tags
strategy
/// RELATED
Strategy4d ago
Artemis III aims for 'late 2027' for Earth orbit demonstration
NASA targets late 2027 for Artemis III Earth orbit demonstration of SpaceX and Blue Origin landers, setting up a 2028 lunar landing attempt with interoperability testing in between.
Models1d ago
Using "underdrawings" for accurate text and numbers
A simple "underdrawings" technique outperforms Gemini 3.0 Pro and ChatGPT-Images-2 at rendering accurate text and numbers in AI-generated images.