Sid, you're with us at NYSE this year. You've had some big announcements and talk to us through that. We had a big announcement yesterday, our first publicly announced heterogeneous computing deployment. 

So you have white coders, coding enthusiasts, all looking to use coding tools, and they want this highly interactive experience when they work with coding tools, which is leading to this need for extremely fast tokens. 

And NVIDIA GPUs, or any GPU for that matter, doesn't have the ability to, on its own, generate those fast tokens. 

So you pair it with a solution like ours, an XPU, that is very good at low latency decode, and create a whole new class of user experience for anyone who wants to do coding.