According to CoinMeta, NVIDIA announced that its investment in the Groq chip has officially entered full mass production. Last December, NVIDIA spent approximately $20 billion to obtain the technology licensing for Groq and poached core team members including founder Jonathan Ross. Eight months later, the first hardware Groq 3 LPX was officially launched; this system is specifically designed to accelerate large model inference and is equipped with 256 Groq 3 LPU chips. During testing, the Groq 3 LPX was capable of outputting 3,431 token per second, which far exceeds the performance of the fastest API on the market, which is about 870 per second. Nebi.us will become the first cloud provider to adopt the Groq 3 LPX, and Groq will also collaborate with Dell for deployment.