Elon Musk makes $43 Billion offer for private buyout of Twitter

Not necessarily, Telsa has moved into building their own in-house processing chips called Dojo, which is a wafer style chip that stacks 25 high powered processors that acts like a single processor and is expected to be a powerful solution for AI training, eliminating Telsa's need to buy and run NVidia chips.

https://www.tomshardware.com/tech-i...-a-serious-processor-for-serious-ai-workloads

NVIDIA is now moving on from H100 chips to Blackwell which will be much more powerful and roughly 30% faster than the H100 ( from what i have read )

Currently, the Dojo Wafer runs about 25% of H100 capacity and requires liquid cooling. Thats not enough for his Dojo Supercomputer which will be used exclusively for AI training

And when i read about the AI training and think of what it will have to be able to do....i dont know that he pulls this off. This Dojo SC will be responsible for spitting out the data that will be used in autonomous vehicles. ( self driving )

The magnitude of this endeavor is really hard for me to wrap my brain around.