Hi Dotikers!

Merry Christmas everyone! While you're unwrapping your presents this morning, Jensen Huang decided to get his own: Groq, for the modest sum of 20 billion dollars. It's the largest acquisition in Nvidia's history, and it lands on December 24th. You can't say they lack a sense of timing.

Yesterday, we discussed AI capable of thinking six moves ahead to solve Zelda puzzles. Nvidia, clearly, is playing the same game but on the hardware chessboard. Because Groq is no ordinary player. Founded in 2016 by Jonathan Ross, the man who invented Google's TPU, the California startup developed a revolutionary chip: the LPU, Language Processing Unit. This architecture is specifically designed for inference, the phase where AI models move from training to real world deployment. The announced result: ten times faster and ten times more energy efficient than traditional GPUs.

Nvidia already holds 92% of the data center GPU market. With this acquisition, the California giant also locks down the inference segment, precisely where AMD and Intel were trying to carve out a space. The message is clear: no mercy.

What's fascinating is the deal structure. Officially, it's a non exclusive licensing agreement. Groq remains independent, GroqCloud keeps running, but Jonathan Ross and his team join Nvidia. They're paying 20 billion for technology and brains, not for a company. Financial artistry at its finest.

In September, Groq was valued at 6.9 billion. Three months later, Nvidia signs a check three times larger. When you have 60 billion in cash sitting around, you might as well treat yourself to nice gifts.

G.