“we said we thought that would begin to convert to revenue in the second half of the year. And we saw that again, the first piece of that in Q3, and we expect to see a bit of it again in Q4.”

In context

“ith some of these web scales and enterprise is growing the rates are calling out, why isn't this sort of growing faster than even the 8% on the comp here? And then just following up on Simon's question before, as my follow-up. Can we just an update as to the exposure between data center versus campus within networking? And how we should think about the price per port kind of playing out here as we're 7, 8 years now down the road? Thanks guys. Chuck Robbins : On the first one, I think the short answer is that most of this AI infrastructure has not begun to flow through revenue yet. Scott, I don't know if you want to --. Scott Herren : That's right. We said all along that we thought the -- when we put the $1 billion target out for the year, that was a sales target, not a revenue target, but we said we thought that would begin to convert to revenue in the second half of the year. And we saw that again, the first piece of that in Q3, and we expect to see a bit of it again in Q4. Chuck Robbins : And I think on the data center versus campus, clearly, the data center with a higher speed, you're going to see a higher per port cost, but you are going to see much higher volumes in the campus. So I think they sort of normalize each other out over time. Sami Badri : Thank you Jim. Michelle, we can move to the next analyst. Operator: Thank you. David Vogt with UBS. Andrew Spinola: This is Andrew for David. I wanted to ask a question about the texture of those $600 million of AI orders. You mentioned that they tend to be nonlinear and they're very large customers. I'm wondering if you're starting to see bigger orders is a large component of that $600 million and with the customers where you're having the most success? Or are you really seeing sort of a diffuse kind of larg”