Tim Beyers

Okay.

So you both agree that TSMC is at the head of the AI value chain here.

Last question on this, and then we're going to move to our second segment.

Because, Jason, you mentioned something interesting, and I had forgotten about this, even though it wasn't that long ago, the way DeepSeat got around CUDA.

So, again, CUDA is the programming language.

It's what makes the NVIDIA solution a full-stack solution.

They give you the software to run the code that runs on the GPUs that makes AI happen.

But what DeepSeek did is they dipped into assembly language.

Like you said, they went underneath CUDA, and then they reprogrammed their GPUs to run without having to use CUDA.

They kind of got around NVIDIA.

So, give me your...

Prediction, starting with you, Travis, how easy is it going to be, or will we see it in 2026, that more companies get around CUDA such that they choose to say, you know what, I don't need NVIDIA anymore?

What is the traction, put another way, that non-NVIDIA GPUs and CPUs get in 2026?

Jason, what do you think?

What's your guess, your prediction?

Okay.

2026, not the year.

Up next, we give some honor to Dr. Martin Luther King Jr., and we talk about the best corporate citizens.

You're listening to Motley Fool Money.

All right, welcome back to Motley Fool Money.