$CEREBRAS - Cerebras recently IPO'd at $63B market cap with 15-20x speed advantage over GPUs for AI inference across all model types. Signed major deals with OpenAI ($20B+) and AWS, with backlog north of $20B. Company is scaling manufacturing 10x this year to meet surging demand as AI models became useful enough for daily work in 2025.
$OPENAI - OpenAI recognized the critical importance of fast inference and closed a $20B+ deal with Cerebras in just 4.5 weeks, demonstrating exceptional execution speed and focus on performance. Company is working to keep up with massive demand.
$COGNITION - Cognition (maker of Devin) is experiencing exceptional growth rates never seen before, demonstrating the power of fast AI inference. Company acquired Windsurf over a weekend, showing aggressive execution.
$CURSOR - Cursor is experiencing unprecedented growth rates in the AI coding space, demonstrating the market demand for fast AI-powered development tools.
$AIINFRA - Fast AI inference is becoming fundamentally important as AI integrates into daily work. Speed advantage of 15-20x creates entirely new business models beyond incremental improvements, similar to how fast internet transformed Netflix from DVD delivery to movie studio. Market for slow inference will be zero.
$OPENSOURCE - Open source AI ecosystem has kept the market vibrant and enabled a generation of companies to do impressive things. It pushes closed-source companies to innovate and has allowed creativity to flourish, particularly with techniques from Chinese makers.
Bearish:
$NVDA - GPUs are 15-20x slower than Cerebras for AI inference across all model types. As AI becomes integrated into daily work, speed is fundamental and the market for slow inference will be zero, creating competitive pressure on GPU-based solutions.