You know, it's about the full stack. And the easiest way for our partners to achieve that profitability and to take advantage of all the incentives is really to think about selling that full stack, positioning the full stack, which, by the way, is what the customers need anyway, as you are so well aware. 

Most of the infrastructure needs a refresh. So when we're talking about that whole stack, whether it's secure AI factory with NVIDIA, that type of thing, it's posturing not only the compute, but the data center networking, the security, the observability. 

not only at the core, but all the way out to the edge. We talk about unified edge where we bring all of that capability into anything that you might be doing at the edge, which of course is where the inferencing is headed.