Gary Mobley: Thank you, both.
Operator: Thank you. We go next now to Jay Vleeschhouwer at Griffin Securities.
Jay Vleeschhouwer: Thank you. Good evening. For my first question, I’d like to ask a variant of the EDA market environment question. So, on the one hand, what are you seeing in terms of unscheduled new business, that is to say, intra-contract new or expansion business that could be construed positively? On the other hand, how concerned are you about the evident deceleration of semi R&D growth? It’s still reasonably good, better than four or five years ago, but so much lower than it’s been. A lot of that can be attributable to Intel. But still, how are you thinking about those two different dynamics? And then I’ll ask a follow-up.
Anirudh Devgan: Yeah, hi, Jay. Good question. I mean, we are watching it carefully. Like you said — like we said earlier, I mean, design activity is still strong. But of course, the macro environment is challenging. It’s like natural — even though the customers realize that they need to invest in R&D for the future, if the revenue is impacted because of macro situations, that decisions become a lot more prudent. But — so this is just natural business process. But in general, still the large customers in the big segment, they’re all investing in R&D, design activity is still strong. And then we just have to see — we had a good Q3 in terms of bookings, like we mentioned. So, we’ll see what happens in Q4 and that will also give us a better idea going forward.
John Wall: Jay, as you know, a lot of our customers come back and purchase add-ons during the course of their baseline contracts and with the teams releasing significantly new business — new products from the different R&D groups that customers have an intent to come back and keep purchasing. So, they don’t wait — when we launch AI tools, they don’t wait for the baseline renewal to come up or to expire to purchase them, they’ll purchase add-ons and they’ll purchase a few licenses and then hopefully proliferate more on the contract renewal. And as you know, we have a lot of contracts that come up for renewal in Q4.
Jay Vleeschhouwer: Understood. For a follow-up, I’ll ask about some interesting Cadence management comments at last month’s Cadence Live event up in Boston. So, there was an interesting comment about the role of AI as “derisking schedules” in addition to the design exploration use case. And what’s interesting there is historically, schedule risk or completion risk has to do more towards the back end of the process, for example, physical verification. So, to the extent that more of that risk mitigation moves up earlier in the process, do you think that there will be a spending share shift within the totality of EDA spend, perhaps some from the back end, more towards the front end where you play with a lot of your tools?
Anirudh Devgan: Hey, Jay, I would say that it should lead to more design activity if we are able to reduce risk in the design process. I mean, as you know, this is the history of EDA, history of automation, even for the last 20, 30 years. I remember in the old days in the ’90s, we would take like five years and 500 engineers to design some big chip. And now that takes six to 12 months and maybe engineers. So that’s like 100 times more efficient than 25 years ago. And I think AI can, as you know, provide the next generation, next level of improvement in productivity and risk mitigation. I mean, part of it is also risk mitigation. So then I think it should lead to more design activity, especially by the system company. Now the shift of front-end to back-end, I mean I think back-end is still a complicated process.
And so, I think even though some of the things can be pulled upfront using AI or using hardware platforms, I still think the back-end design process requires a lot of work. So I would expect it affects all of them. And then, the other thing we are trying to do on the front-end, as you may have noticed, is really incorporate LLMs like our partnership with Renesas, because a lot of the front-end process has been less formal. The back-end process, especially once we have RTL, then we go to gates, we go to GDS, it’s a very formal process, very structured process. But the front-end of the process, especially verification and specification, has been less formalized. And I think AI and LLM can help formalize that which definitely, like you said, can minimize the risk.
But I think activity should still be strong in both front-end and back-end. And our goal anyway is to make the design easier, so more customers and more people can do them.
Jay Vleeschhouwer: Thank you, Anirudh.
Operator: Thank you. We go next now to Jason Celino at KeyBanc Capital Markets.
Jason Celino: Great. Thanks for taking my question. Maybe first for John, on the Q4 guide. Apologies for asking this again, but folks might be wondering tomorrow, I guess, why aren’t we seeing more upside to the guide for fourth quarter? Because where might be some conservatism or what way will you be overlooking in terms of the setup?
John Wall: Yeah, Jason, as you know, your question probably emanates from the fact that we beat by $23 million in Q3 and raised by $10 million, but that was mainly due to a prudent guide for Q3 with respect to certain hardware installations. I think overall, we’ve taken the quarter up — or the year up by $10 million at the midpoint. But I think there’s — because it’s expected to be a strong bookings quarter and a particularly strong quarter for our IP Silicon Solutions group that — I mean, they’re having — they’re going to have an excellent Q4 that we’re expecting that all year. I think if there’s upside, it’ll probably come from that group.
Jason Celino: Okay. No, that’s fair. And then just my quick follow-up on backlog. I know you’ve got some weird comps because of the hardware stuff. But when might we see like year-over-year growth again? Or I guess if we stripped out the hardware-related backlog, I don’t know if there’s any way to share what type of growth you might be seeing?
John Wall: Yeah. I think just to give you a bit of color on that, I think if you recall, at the end of last year, our backlog included about 28 weeks of lead time on hardware. I think we’re down to an eight to 10 week range now on lead time for hardware. So, of course, we’ve eaten some backlog as a result of that. But I think we troughed out essentially in the middle of the year. We’re expecting the second half to be stronger for contract renewals, because the number of contracts expired in the second half. In Q3, you saw backlog starting to tick back up again. We’d expect it to tick back up again in Q4 because we have a strong bookings quarter, or we’re expecting a strong bookings quarter. The one I’d look for really is the annual, the kind of CRPO is the one I track, because I’m looking at the annual value.
