Groq’s $6.9 B AI Chip Surge: Inside the Inference Revolution
Groq’s History and Mission Founded in 2016 in Silicon Valley, Groq was born when a team of engineers from Google’s Tensor Processing Unit project left to pursue a bold idea en.wikipedia.org. CEO Jonathan Ross, one of the TPU’s designers, believed that as AI models grew, the real bottleneck would shift to inference – the process of running those trained models in real time medium.com medium.com. Groq’s core mission became delivering fast, affordable AI inference at scale, encapsulated in Ross’s vision to “drive the cost of inference toward zero” medium.com. Groq’s philosophy is rooted in first-principles thinking. Rather than repurpose existing