Andrew Feldman

It's why they can't be fast.

It's why they can't be fast.

It's why they just spent $20 billion buying Grok.

It's why they just spent $20 billion buying Grok.

It's because they didn't have an answer for fast inference.

It's because they didn't have an answer for fast inference.

Fast inference needs fast access to memory and the GPU doesn't have it.

Fast inference needs fast access to memory and the GPU doesn't have it.

So these are things we're constantly thinking of.

So these are things we're constantly thinking of.

I think everybody knew that the demand would increase.

I think everybody knew that the demand would increase.

And this is true among the major memory makers.

And this is true among the major memory makers.

People get a little scared.

People get a little scared.

And what happens is they place a full year's worth of demand.

And what happens is they place a full year's worth of demand.

And they get the wrong answer back, which is, we don't exactly know when you can have it.

And they get the wrong answer back, which is, we don't exactly know when you can have it.