Good afternoon, everybody. We're gonna get started with the next presentation. We're happy to have Cadence here. We have the CFO, John Wall, as well as head of investor relations, Richard Gu. I think before we start, Richard is gonna read the obligatory disclosure, and then we'll hop into some Q&A. And if you have any questions during the process, I'll look up every now and again. Just raise your hand, and we'll run a microphone over to you. Wait to ask the question until you get the mic. So with that, I'll hand it over to Richard. Good day, everyone. So today's discussion will contain forward-looking statements and will make use of certain non-GAAP financial measures. Please see the company's, our most recent 10-K, 10-Q, and website for a discussion of risk factors and, for our risk factor, and our use of non-GAAP financial measures. Thank you. Glad you read that, not me. Okay, thank you for coming. So, John, so maybe we can begin with some of the short-term themes, and then we'll go into some longer-term structural questions. So starting with the company's vision over the next 12 months or so, or the visibility into that, could you perhaps give a brief overview of the pipeline strength, diversified streams of revenue, and how to think about the main drivers of growth, potentially on the operating margin side of things? Yep, fair point, Ross. Thanks, and thanks for having us here. The I think the way to think about Cadence is that Cadence, what we realized is our strength is in computational software. And if you take computational software and you apply it to silicon, that's EDA. When you apply it to you know a phone or a building or a car, that's system analysis, and then when you apply computational software to data, that's AI. And if you look at Cadence, I mean, under the umbrella of Cadence, there's probably five businesses. I think one of the engineers retired there last year, and he told me it was the best five companies he'd ever worked for. But we have a I mean, traditionally, we have an analog business, analog design business. That's performing really strongly. I mean, we're a big portion of the market. We're probably 80% plus of the, of the market on analog. And that custom IC team, you know, there's, we have some of our biggest franchises of Cadence there. Like our-- if you have a 100-person analog design house, they're probably using a 100 licenses of, of Virtuoso. On the digital side, we've made great inroads over the last kind of decade or so on, on digital. I think on digital we're, you know, close, if not at 50/50 with Synopsys now on the digital side on, on EDA. And as a result of our strength in analog, I think we're the largest core EDA company. But... And then, of course, there's verification. Verification has been a really strong grower for us, above average grower. That business has done really, really, really well. The challenge in verification has been a secular one. Like, Jensen Huang believes that, you know, 100 billion transistor design right now is going to have to be a 1 trillion transistor design in by 2030. And like if you think of the growth and complexity as N, the verification challenge is two to the power of N. But, so we've just seen huge demand for, say, our emulation systems on the verification side. We have a whole suite of tools in our verification flow. The IP business has done well. We, we picked our spots in IP. We don't have the biggest IP business, but we like to have a differentiated IP portfolio. And then, of course, on the system analysis side, system design analysis has been a fast grower for us. It's about 12% of our business now, probably around $500 million annual value. But, it's an area where I, I think we have the most headroom for growth 'cause we're newer to that area of the business. But it sits really in that sweet spot for us because we have such a strength in computational software and in simulation, and that comes together in the system analysis space. On the analog side of things, how is the consolidation of that industry? Does that matter to you? It's seemingly largely done at this point versus the complexity increasing. So does it matter to Cadence if it's, you know, one big company doing all of it or 10 smaller companies doing the same thing? Does that matter at all? No. I think what matters to us mostly is the number of design starts and project starts. I mean, the more project starts where people are customizing their own silicon, the better it is for us and other companies in our industry. Do you find that fewer companies that have the wallet to do more of the design starts is the output of this industry consolidating? Or when it was a number of smaller companies that were all trying to do the same thing and, you know, they all wouldn't win- Right. but, which of those tend to be better for you? Oh, again, we're agnostic. I mean, we help everybody. We try to partner with everyone. But now, of course, as you move down the process nodes, the more advanced nodes, you always have the deep pocket customers, I think, that lead the way there, but others tend to follow over time. I think what's more important is the amount of volume. But, I mean, if you're-- you might be spending a lot for your design activity in developing your own chip, but if you have the volume requirements, you know, the price per unit is not prohibitive. So I would imagine with the rise in lagging edge capabilities and necessity of lagging edge nodes for a lot of the analog capabilities, that's gonna be a big positive for you guys as well. Again, I think, you know, it's one of those things that the more people that are taking it upon themselves to do design projects, it's just, it's, it's really good for us. That it's like there's an arms race going on, and we're like your friendly arms dealer. But- Gotcha. What was the— Just to follow up to the, another part of your first answer, the digital side, where you talked about how you penetrated that and you're 50/50 with the Synopsys guys. Now, what was the key?... to penetrating that, and how— what was the trajectory from where to the 50/50 at this point? Oh, in fairness, that like, if I go back through history, Cadence was formed back in, I think, 1988, but by Alberto Sangiovanni-Vincentelli. He's the-- He's on our board today. He's a professor at Berkeley. And when he formed Cadence, it was an analog EDA design house that analog EDA company that. And I think he helped, I think, in history, the, I think he helped Aart with develop Synopsys. I think Aart had this idea to create a digital version of Cadence, and I think Alberto. He consulted with Alberto on that at the time, and Alberto served on both boards for a number of years. I remember teasing Alberto once, saying that, "Oh, Synopsys is bigger than us. Did you, did you choose to stay on the wrong one?" And he said, "No, no, no," because he told me that his expertise was in EDA, and he said that our destiny was always to be the largest in core EDA. And he figures that the destiny was meant to be our destiny. He was so confident in that destiny because he said analog to digital was like orchestra to rock band. And he said orchestra players could play in a rock band, but it would be tougher for rock band players to come the other way. And he felt we'd have to do something terribly wrong to lose our strength in analog design, and that's kind of how it's played out. But he felt that it was only a matter of time before we got to 50/50 on digital, but we needed the right leader for that. We got that right leader with Anirudh Devgan, who's our current CEO. But we picked up Anirudh about 12 years ago. Synopsys acquired Magma, and we, Anirudh left Magma and came and joined Cadence. So it was a very, very lucky moment for us. And Anirudh came in, and I think he taught us how to do digital. But, because he took, he looked at our digital tools and said, "You know, you're trying to be as good as Synopsys, and as good as is not good enough." That, you know, it needed to be multiple times better to convince someone to change. So he scrapped how we did digital design and revamped our tools. And actually, the methodologies that he used to revamp our digital tools and the techniques that were used there were the basis of our simulation capability that we're applying to system analysis today. But so again, Alberto was right that, you know, it was only a matter of time before we got to 50/50, where there's still some big customers that do a lot of their digital design with Synopsys. But we think, outside of maybe one or two big players, we're at least 50%, if not more, of the digital market. Great. Well, let me move on to the next question on the list that I sent to you guys. Is the semi industry deals with an acceleration to more complex and lower nodes, geometric nodes. How is Cadence placing itself with regards to both EDA tools and IP products to help with the tighter supply of engineers and the growing demand for design starts? Well, I think, like, if you, if you look at our customer spend, let's say our semi customers, that for every $10 million they're spending on R&D, they're probably spending $9 million on people and $1 million on tools that they give those engineers to do their jobs. But, now, over a period of time, of course, that will double, that, it will go from $10 million spend to $20 million spend, and, and if they don't adopt AI tools, it'll probably stick with... If, If, if it's nine to one, it'll be 18 to two. But what we're trying to convince our customers of is that with the capabilities of our AI tools, is that you, you'll probably get the benefit or the productivity that would take you $20 million to do without AI, for maybe $16 million. But only if you spend 25% of that $16 million on tools. And we think it's inevitable that the industry goes that way because the growth and complexity of design is far outpacing the growth in human population, never mind the growth in the population of engineers or the engineering community. So if we are going to get to, like, Anirudh's vision of, like, 100 billion to 1 trillion transistors by 2030 on a single chip, on a 1-inch by 1-inch chip, it's like 30, 40. I mean, the amount of complexity is just enormous, and you're not gonna get there with just brute force and more engineers. It'll have to be better technology and better AI capability. I think, Richard, you have a great example, don't you, of, of our use of AI technology. Do you want to talk to that? Sure, John. You know, AI definitely is a very exciting area, right? That's one of the one of the technology that can help us bridge that labor shortage gap, which we're all facing. One example we, we had done as a company is we have this product called Cerebrus, okay? It's a digital AI tool, like optimization tool, sitting on top of the the digital full flow, from synthesis to place and routing to the final sign-off. And we're working with this big automotive semiconductor company in Japan, Renesas, and they had this design, which is dealing with 17 variables. When you think about it, you know, it's inconceivable for a human being to explore exhaustively the entire design space because it would require 4 million runs, and it'll take hundreds of engineers months or even years to, you know, exhaust the space. With Cerebrus, they were able to not only do it within weeks, but with a much better PPA. PPA represents power, performance, and area, which is the litmus test and criteria in terms of how good performance, how good the product and outcome is. So... then, within a much shorter period of time, they were able to achieve much better results with a lot less engineers. ... So I think it's one example of that AI, generative AI, was able to kind of bring tremendous benefits to companies and customers, completely revolutionizing the entire space. And one of the great things is also it has this transfer kind of learning process using reinforcement learning, where you're parlaying the learning from the previous run to the sequential next runs. And then by leveraging the compute power of many CPUs, you're able to kind of dramatically shrink the time to market. This is one example how of how these great products are bringing benefits to our universe. How do you... That's the perfect segue into the whole AI theme. How do you guys look at the contribution to Cadence's top line from AI? You know, whether it's Cerebrus or in a more general AI suite and operating margin expansion. How do you kind of fold that in and give metrics for people to monitor the progress? Well, generally, I mean, it's a very, very innovative company. We spend 35% of our revenue every year on R&D. And, you know, more than 50% of that R&D is probably not contributing a dime of revenue this year. I mean, it's future projects and developing tools for the future. You know, a lot of our revenue is coming from tools that were developed for previous process nodes, where the R&D cost is already sunk up front. So it's a very, very profitable company, very cash flow rich company. The opportunities we have to take that cash flow, it's almost like a virtuous cycle. You know, the benefits we get from the profitability generates the cash flow for us to continue to invest in the next generation of technology. But, but our model is essentially that we aim to achieve double-digit revenue growth. AI contributes to that. That's just the latest innovation, and we're always innovating. But, double-digit revenue growth with improved margin profiles should end up with kind of, you know, low teens operating income growth. But, and then we use 50% of our free cash flow to buy back shares, and we're very disciplined in our share-based comp. But, so the... you should see the share count decline over time. What we've seen over the last 5 years is that when you add all those things together with the declining share count, that, you know, double-digit revenue growth has resulted in, you know, low- to mid-teens operating income growth, which has resulted in high teens, low 20% EPS growth. You know, that model works. I mean, when Anirudh and I started working closely together, back in kind of 2016, 2017 timeframe, that we found that the R&D engineers, we wanted them to be more disciplined in terms of their approach. We wanted to be more commercially aware. So in an effort to change the culture there, we made the R&D teams compete for investment dollars. But, so we had an investment pool, and then we'd make them compete for the investment pool of dollars to make them more commercially aware. They're dangerous now. They come looking for investment dollars, and they're showing us the press release, what the press release will look like, if you give me the dollars to do this. But I think they're much better at focusing, like thinking like an investor, thinking like an owner, and choosing projects or allocating the investment dollars to projects that they think will generate a return for investment. Perhaps as a bit of an aside, but you just mentioned about being disciplined on the stock-based comp side of things. For software companies and semiconductor companies, you know, I'm old enough to remember when people found that religion and backdating and all of that 15 years ago, but they seem to have seemingly lost that discipline. Do you have any trouble on the hiring front, given the competition for engineers these days and the compensation programs with stock-based comp that so many of the other folks in the valley are offering? No, not at all. I mean, the benefit that we have, of course, we've registered high in the rankings of the great places to work and very, very consistently, I think for the last 7, 8 years. There's, you know, the team at Cadence has done tremendous work to make it a great place to work. And the, I think people enjoy the work that they do at Cadence. And for the engineering community itself and the new college graduates that come in, one of the things that they tell us is that it- when they go to some other companies, that there's a specific discipline. That, at Cadence, there's like 5 companies under the umbrella of Cadence. You get a broad exper- a broad array of experiences that's operating at Cadence. If, if you want to try your hand at different engineering disciplines, there's the opportunity to do that at Cadence, and I think it's helped us attract the best talent. One thing, going back to the second part of the last question I asked, you also mentioned that double-digit revenue growth was the target. I believe that used to be high single-digit revenue growth. What's causing the step up? So the growth area that we've had, I mean, so your EDA to traditional EDA business is probably a high single digit to low double digit grower. But functional verification has been on fire the last few years, particularly the emulation systems, hardware systems. But that's been an above average grower for us. We've discovered that our computational software capability, when it's applied to, like, physical things like buildings and cars, and that it opens up the system analysis space for us, and we have tremendous simulation capability. The system design analysis segment of the business, now 12%, but that was 12% of Cadence's revenue, but that was very, very low before. So that's been growing really, really fast. And then on the IP side, the IP business has done quite well. We've been picking our spots there in IP. We want to have a differentiated portfolio of IP, and we aim to do profitable and sustainable revenue growth in IP. Switching gears a little bit, I think it's noteworthy that Cadence now holds about 45% or generates about 45% of their revenues from non-semi companies. How exactly are you positioned in regarding those customers versus the semi customers, and how does the activity compare to the traditional semiconductor base? Yeah, so I'll take that, Ross. What's interesting is the semi, in our view, the semi-traditional semiconductor companies, they're increasingly converging with the system companies because you are seeing semi companies trying to get into become a system company, develop their own softwares, and system companies are trying to develop their own custom silicons in order to support their software stack. So Cadence stands to benefit from that mega trend, and we feel very good about our positioning in that space. Give an example, if you look at the hyperscalers, right? They're all going in and try to develop their custom design silicon in order to be tightly integrated with a software stack. That brings tremendous benefit in terms of not only the best performing kind of, you know, silicon, but also the opportunity to really control their schedule, right? Not to mention the ability to save a lot of cost, right? Given the price of the silicon going up. So, this is an irreversible trend for, from our perspective, and it's at a very early innings of development. Also, in addition to that, I think it's good to keep in mind that the traditional semi companies are also doing a ton of design activities, like as John just mentioned. If you look at the proliferation of the silicon is everywhere. Hyperscalers are starting point, mobile phone, the waves, and then now you're seeing a lot of automotive kind of actions in that front. So we feel like, as a company that has a very strong portfolio, starting with EDA, now expanding into SD&A, System Design Analysis, which is highly synergistic with our EDA business, but much more accretive in terms of margin. And we have this unique packaging and, packaging and PCB business, which is the glue that ties together those two big pieces of, you know, kind of businesses. That really sets up well for us to really tap into that growth opportunity and potentials. So if you have that, that hyperscaler doing their own silicon trend is, I agree, is irreversible. Some of them have a partnership with the silicon vendor, you know, the ASIC division of, say, a Broadcom or a Marvell or something like that. When you have any of the cloud or the hyperscaler guys going down that road, do they work with you, or is it the, in my example, the Marvells and the Broadcoms that actually work with you, or is it both? Oh, it's, it's both. Yeah, we work with everybody. I mean, ultimately, what we'd love to get to is that, you know, chip design becomes a middle school project at some point. Right. Make it as easy as possible. The more democratized it is, and the more easy it is, the easier it is for people to design their own silicon, custom silicon, the better it is for us. But, and I know that sounds farcical and far-fetched and everything, but, you know, I would imagine if we said something 20 years ago that you'd be coding as a project in middle school- Right. People would have thought that that's far-fetched. But the ability and the benefit of using AI tools should hopefully make this easier going forward. And like I say, the more, the more companies there are, the more design starts there are, the more companies that are designing their own silicon, I think the better it is for Cadence. But I think, yeah, the complexity side of the equation is why it doesn't- Yeah. Democratize it down as far, and the AI example is the perfect one. Yeah. That complexity goes off the charts with, like, the 17 variables that you talked about, Richard. So I would assume the design intensity and the necessity of your offerings in AI and a general theme goes up massively. So that has to be... You guys have to be a great play on that theme just to allow these companies to have the capabilities to address that market, let alone do so efficiently and profitably. Is that your view? Oh, yeah, absolutely. I mean, when you look at some of the biggest companies in the world, I mean, just like a Google search engine is complex, but people understand it. But, I mean, if you looked at the complexity of what's going on behind it to, to, to give you the solution as quickly as it, as it does, the, people might be afraid of that complexity. I mean, when you look at Cadence, I think a lot of investors probably stayed away from EDA because it, it, it was more in the... It was easy to put into the too hard to understand bucket. But I think it's, it's, you know, we've really evolved into being a computational software company. And we're just trying to make it as easy as possible for multiple users to use our capability. So changing topics slightly, just wanted to see if you could share your vision on your IP portfolio, and how does the new acquisition of the Rambus IP tie into your strategy? Yeah, sure. We have a very unique IP portfolio, right? IP is an important business for us. It is right now about 11%-12% of our revenue and continue to grow. Our IP strategy and philosophy is our IP, meaning that we're not trying to be everything to everyone. We're focused on developing and driving the value accretive, high margin, you know, really essential IP. So we have design IP, and we have another IP business called Tensilica, which is a processor IP. These are highly sticky and highly profitable kind of IP business. With Ross, to your question of our acquisition of Rambus IP business, it's a very strategic move for us. It's really driven by a couple of major considerations. Number one is it's a commercially available IP, right? So you don't have to throw bodies at developing our IP and try to make it work with the customers. The second major point is, that IP has HBM3, high bandwidth memory three, which is very attuned to support the AI use case and portfolios. That's why we feel like, you know, that, that's a very, very attractive kind of IP portfolio we, we, we have in our, on, on our wings. So we would feel like, you know, with that, to help us continue to drive that, that book of business for us. Got it. On the verification side of things, could you talk a little bit about the hardware demand driven by the increasing need for verification? When do you see it slowing, if at all, and is it becoming more correlated to the EDA sales cycle, or is it still kind of a cyclical business that's simply catching up to what was a lackluster demand environment throughout COVID? Oh, it's verification is one of those problems that people don't solve, but they manage effectively. I mean, the growth and complexity of design is so much that it's hard for them to keep pace. We do think there's a secular trend in demand for things like our emulation capacity. We've had probably eight quarters in a row now where demand has outstripped our ability to supply. But, I mean, in saying that, demand for, say, hardware systems as part of our functional verification suite, that can be seasonal to a certain extent. I mean, Q4 is normally bigger than Q2, for example, but our Q2 was a record Q2 for us, in terms of demand and sales. But the whole verification suite and the flow that we've put together, and that team have put together, has served us, our customers very, very well. We believe it's a, like I say, a secular trend. And yeah, I guess that's what we all to say there. Is there anything else you'd ask? I mean, you covered everything. Okay, fair enough. Yeah. Good answer. Actually, I have a question from the audience over there, if we could get the microphone. Sorry, to go back to AI a little bit, but the kind of two interrelated questions I have are: What data rights do you have with respect to your customers as it relates to generative AI? Obviously, these LLMs have to overlay on top of data, and I don't know if your customers are allowing you to utilize data in a way that helps all, all of the customers, if that makes any sense, or just a specific customer. And then the second question relates to, do you think directionally that generative AI will basically democratize and provide access to semiconductor design by individuals who maybe don't have a double E? I mean, do we get to that point, I guess, long-term in the industry, where engineers of different types might you might have new customers that you didn't have before? Yeah, so I think I'll take the second part to your question. I'll hand it over to Richard on the first part on data rights. In relation to generative AI making it easier, I guess, more user-friendly to develop custom silicon, I think that would be that's a clear benefit. That's anything that makes it easier for people to design custom silicon is a benefit to us. And there's I think it does a lot of the heavy lifting for the more talented engineers to focus on the higher value add activities. But so I think it'll help there. Now how user-friendly it gets over time is yet to be known. But when I talk to Anirudh and he thinks that AI, I mean, it's very, very real. It's like the internet. There will be waves. Right now you're seeing an infrastructure wave. But we think there's many, many applications that we've yet to see for AI. On your data rights thing, I don't think we're as dependent on the customer's data in terms of how our tools work. But do you want to talk about that, Richard? Sure. Sure. Sure. Yeah, I'll just give you a few points. From a data standpoint, obviously, we're very judicious in terms of use in terms of making sure every customer's data is only available to that specific customer, right? Because data is the key, and it's private, and it's so important. It's like, it's like the lifeblood to drive the betterment of the AI, AI tools. We have this platform, like big data platform called JedAI, which really kind of stitch together all the five APIs we developed that covers the entire gamut of the life cycle for chip development, from the design, to verification, to packaging, all the way through system analysis. So the great thing about that is the AI products are only as good as the data you can generate and how much visibility you have in there, right? Once you have the entire big kind of platform and portfolio that covers the entire life cycle, customers will have visibility in there, and they'll help them continue to fine-tune and make the AI kind of, you know, APIs better. And another thing is, those AI tools, our JedAI platform is built using the open source data like Spark and Hadoop. So what it means is that customer can use their own APIs and kind of, you know, latching onto that data platform, data lake, to help them continue to make their products and design better. Any other questions from the audience? We have one in the back. Hi. Ian Murray. I was kind of struck by the percent of revenue that you spend on R&D, and certainly as tech investors, we understand that product is the lifeblood of the company. And you seem to have, you know, the five different companies, as you said. But how is that spent? Is the analog... doesn't require a lot of R&D currently, given its status. Certainly, AI would, but that's really a high number, a high amount of real dollars. So could you delve into a little bit how that money is spent? Yeah, sure. Yeah, in terms of the allocation of the R&D dollars, I mean, like I say, 50% of it is to new projects and to the development of tools that are not generating any revenue today. But when I look at the spread of R&D across the business, there's IP is probably our least profitable business under the umbrella of Cadence, so more of the R&D dollars would go there than, say, analog is a very profitable business for us. But when you look at, you know, chasing Moore's Law and the amount of innovation that requires to keep moving down through the process nodes, it does take that amount of investment to keep up. One thing I worked on recently was I was trying to do a. I'm a value guy, really, in a growth company, a value guy stuck in a growth company. But, and when I'm looking at things like our buyback policy for repurchasing shares, I'm always looking at the net present value of future cash flows. So we do this model that kind of looks out a couple of decades to see what net present value future cash flows look like. And just as an exercise, I wanted to look at what does it look like if we didn't have to chase Moore's Law? If Moore's Law slows down and that, you know, the- and a growth company turns it, morphs into a value company over time, well, then you'd take your 35% that you're spending on R&D would probably decline to something like more normal, like a 5% over time. But, but when you do that, it's quite interesting. Multiples come down, but the cash flow generation and the profit generation, more than makes up for it and the near term, you know, the time that you're able to collect that cash, you end up with pretty similar valuations, whether Moore's Law continues on indefinitely or whether Moore's Law slows down over the next 20 years. But, but yeah, I mean, it's interesting. Anirudh does the allocation of the investment pool of dollars, but there's a... I do a general allocation, and then I tend to tax all of the R&D groups. If you're doing the same thing as you did last year, I expect you to do it more efficiently. So, we take 2% or 3% off of them, create an investment pool for Anirudh to allocate to new projects. And then we just encourage the teams to compete for those, and try to think like owners, in the efforts that they take on. That's when they come to you with press releases. Yes. Yeah, yeah. They're trying to sell us on what we should spend money on. The question over on the right side. Yeah. Could you talk a little bit about China a little bit? How much of your business is there? Have there been any restrictions by the U.S. government selling software there? How far along are the Chinese in their own design software efforts? And if there were restrictions today, cut off to sell software to China, how badly would it cripple their semiconductor industry? Whoa! So, okay, big questions, but a lot to unpack. You have 35 seconds. Okay, so let me unpack a little bit. In terms of China, two of the last three years, 15% of our revenues come out of China. I think this year it's probably gonna be closer to, you know, 16% or 17%. Very strong region for us. There's a lot of design activity in China, a lot of work being done. We are impacted by the export regulations. The biggest impact to us was probably, what was it? 2018, when Huawei got put on the entity list. Since then, there's been some other changes, but they don't have as much impact. That was probably the biggest impact for us, the Huawei thing in 2018. What the government did last year was very thoughtful. I thought that I don't think they got the credit they deserved for the way they went through it. I mean, they introduced some restrictions around gate-all-around technology, which was very helpful for us to understand in advance of getting there. So now we'll create, you know, a version of our Innovus Digital tool that has gate-all-around capability, and we'll create a version that doesn't have it. It's kind of like, you know, a Coke and Diet Coke or Pepsi and Diet. So we'll have a specific tool for the China market, and we can create that. But I think I mean, we're very pleased with the growth of our business in China and with our ability to work with the government restrictions that have been put in place. If there were any further restrictions, I mean, ultimately, you know, it costs the same amount for us to develop our tools and create our technology. But, if we had more restrictions in terms of where we can sell, we would have to increase prices elsewhere. I mean, that's our lever, I guess. Hope that answers most of your questions. Thanks. All right, John and Richard, we are officially out of time, so thank you so much for your views and attending the conference.