Andy Halliday

So anyway, that's cool.

And now they are also launching on EC2, which is their sort of useful, it's called Elastic Cloud.

And what that is, is you can have your website or your application or your operation running on EC2.

And that's a server farm in effect that is scalable, meaning I can rent for right now,

this particular application server, or I can rent this particular AI inference server.

And then, oh, it's Black Friday?

I want to make sure my system keeps up.

So temporarily, I can say, OK, add three more of those.

So I want four operating at this time.

And that's what they mean by elastic.

I can change the number of servers that are being called through the front end of the load balancer for all the incoming traffic that way.

They're launching a new EC2 service that uses NVIDIA GPUs.

Okay, so now you can call using EC2, you know, AI inference that way.

So they have this, you know, that expansion is including the new GB300 Ultra servers from NVIDIA.

So that's like the highest level of GPU memory and compute on AWS.

It's ideal for instance, scale and so on.

So that's on the chip side, on the cloud computing side, they're moving forward to really be a big player in providing inference compute to whoever wants it.

And simultaneously doing major investments with big operations in specified data centers like the Humane one over in the Middle East.

Okay, so now they have released their family of AI models.

I don't know if we go back to 2023, you know, bedrock was supposed to be a really cool possibility.