It's been kind of fun, hasn't it, to see the Tech Titans trade shots back and forth with billions of dollars at stake.

Our audience, they might remember that back when Microsoft first showcased Chat GPT,

Satcha Nodella said he wanted to make Google, quote, dance.

Ever since we've seen a fice year more aggressive Google

on the generative AI front, the company

has now published a scientific paper

that seems to fire back, letting Nodela and Invidia, another giant in the space,

know that not only are they dancing, but they don't even need a partner here.

The paper says essentially that its supercomputer is faster and more efficient than comparable systems from

invidia and that matters because the chips at the compute level is what

drives the products like chat gp t and barred. Large language models

they are becoming so big guys that they cannot be

stored on a single chip. So they have to be split across thousands of chips, which

then have to work together creating a supercomputer. So Google here is saying that it its in-house chip, it's known as the

TENSER Processing Unit or TPU, is being used now for 90% of its work on AI

training and Google says it can be strung together to create better generative

AI than comparable systems from NVIDIA. Now the context here is that more

big tech is designing their own chips like Apple and its M1 series that displaced Intel.

There's also Amazon's custom computing chips for data centers aimed at lower and cost for its

customers.

And guys, this is likely just the start of the AI battle at the chip or the compute level that will drive

these consumer-facing products. At this stage though Wall Street it's worth

...