SpaceX Adopts Nvidia Chips Exclusively as Rivals Target Growing AI Inference Market

SpaceX announced it will exclusively use Nvidia chips moving forward, marking a major commercial milestone for the chipmaker. Despite Nvidia's continued dominance in AI model training through its CUDA platform, industry focus is shifting toward AI inference, where competitors Advanced Micro Devices and Broadcom are capturing ground. Bloomberg Intelligence projects the inference segment will grow at a 32% CAGR to $1.3 trillion by 2032, roughly double the $658 billion training market. AMD has secured two $100 billion deals with OpenAI and Meta Platforms, while Broadcom projects AI revenue to reach $115 billion next fiscal year.