Charles Liang: Oh! Very good question. Indeed, we are adding some more buildings in the Bay Area and most of those will be rental buildings, because our growth is faster than what we can build in a building. So it will be a rental facility. And then I did mention about we are looking for another location, hopefully in a little bit of a lower cost state in North America. So we are planning for building a new campus in — we will decide in the next few months, I believe. But short-term, new buildings in Silicon Valley will be rental property.
Operator: Our next question comes from Nehal Choski, Northland Capital Markets. Your line is open.
Nehal Choski: Yeah. Thank you. You may have already addressed this, but what are your expectations for AI revenue contribution with respect to the midpoint of December quarter guide being up 32% quarter-over-quarter?
David Weigand: Yeah. So I think we are expecting really the same performance, Nehal. We will expect it to be in a range of over 50%.
Nehal Choski: Okay. And can you give a little bit more precise number as far as what the exposure was in the September quarter other than greater than 50%?
David Weigand: That’s — we are giving that approximate figure and that’s our guide.
Charles Liang: Yeah. Basically AI revenue percentage continues to grow, but hopefully in ALCN [ph] consistently with…
Operator: Our next question comes from Aaron Rakers with Wells Fargo. Your line is open.
Aaron Rakers: Yeah. Thanks for taking the question and also congrats on the quarter. I am just curious, going back to the supply side of the discussion, how would you, as you are engaging with customers and thinking about their build-out plans in their datacenter footprints for AI, how would you characterize the evolution of lead times on these higher end GPUs? I mean, relative to what it was maybe 90 days ago, how has that evolved and how are you seeing that evolve into the current quarter?
Charles Liang: Yes. It’s a complicated job. However, because our building-box solution and our global footprint together, we have a huge lead time. We support hundreds of customers. So [Technical Difficulty] indeed, relatively, we are able to take care of the allocation, the lead time, the inventory control, product flow-in, more efficient than our competitors. And we are continually improving in that area.
Aaron Rakers: So, you would say that lead times have improved throughout this course this last quarter and you would expect that, it sounds like to continue improving in the December quarter. It’s kind of a fair characterization.
Charles Liang: Yes. And that’s why our inventory utilization rate will be improving. So, before, I guess, we have about 90 days turnaround time.
Aaron Rakers: Yes.
Charles Liang: And I guess that will be improved when the scale continues to grow and when we continue to…
Aaron Rakers: Yeah.
Charles Liang: … leverage our building-box solution, situation will continue to improve.
Aaron Rakers: Yeah. And then the follow-up question would be, as the market evolves more competitively and you see, I know a prior question on the MI300, you have got the Gaudi Silicon. I guess the way I think about it is these customer deployments are longer cycle. It’s not like these decisions are made in a given quarter. So, as we look to these products, particularly the MI300X ramping, really starting early part of next year and through the course of next year. Are those projects that you are already seeing visibility into and that actually adds another layer of growth to the pipeline? Are those projects that you have already been designed in and you are just waiting for those products to kind of launch to really start to see that incremental revenue for those competitive offerings?
