Welcome to the Synopsys earnings conference call for the second quarter of fiscal year 2017. [Operator Instructions]. Today's call will last 1 hour. Five minutes prior to the end of the call, we will announce the amount of time remaining in the conference. As a reminder, today's call is being recorded. At this time, I would like to turn the conference over to Lisa Ewbank, Vice President of Investor Relations. Please go ahead.

Thank you, Ryan. Good afternoon, everyone. With us today are Aart de Geus, Chairman and Co-CEO of Synopsys; and Trac Pham, Chief Financial Officer. Before we begin, I'd like to remind everyone that during the course of this conference call, Synopsys will discuss forecasts, targets and other forward-looking statements regarding the company and its financial results. While these statements represent our best current judgment about future results and performance as of today, our actual results and performance are subject to many risks and uncertainties that could cause actual results to differ materially from what we expect. In addition to any risks that we highlight during the call, important factors that may affect our future results are described in our most recent SEC reports and today's earnings press release. We will also refer to non-GAAP financial measures. Reconciliations to their most directly comparable GAAP financial measures and supplemental financial information can be found in the 8-K, earnings press release and financial supplement that we released earlier today. All of these items, plus the most recent investor presentation, are available on our website at synopsys.com. In addition, the prepared remarks will be posted on the site at the conclusion of the call. With that, I'll turn the call over to Aart de Geus.

Good afternoon. I'm happy to report that our second quarter results were strong. We delivered revenue of $680 million and non-GAAP earnings per share of $0.88, both at the top of our target ranges. We executed $100 million share buyback for a total of $200 million so far this year. And we-re raising our revenue, non-GAAP earnings per share and operating cash flow guidance for the year. Trac will discuss the financials in more detail. The landscape around us is solid with what feels like a continuing stable outlook for 2017. Our 3 customer groups, semiconductors, systems companies and software developers, continue to invest significantly. This is driven by the dawning age of digital intelligence which is both exciting and evolving quickly. Everything around us is becoming smart, connected but also complex. Mounting software content, whether integrated on a chip or as an app, creates huge functionality, development and security challenges. This, in turn, drives growth opportunities around machine learning, automotive, augmented and virtual reality, networking infrastructure and the soaring need for more cloud-based computation and storage. With the unique portfolio reaching from the roots of silicon all the way up to software, combined with best-in-class global support, Synopsys continues to demonstrate that our vision, strategy, investments and execution are right on the money. This is evidenced in all 3 product groups. Core EDA revenue growth has outpaced competitors over the past years. IP continues its double digit growth and our software quality and security group is scaling to critical mass with excellent revenue expansion. Let me provide some highlights from the quarter beginning with core EDA. Synopsys is a mission-critical partner for the most advanced designs. Our EDA and IP collaboration with TSMC illustrates this. During Q2, we announced full flow certification for TSMC's 12- and 7-nanometer processes as…

Thanks, Aart. Good afternoon, everyone. In Q2, we closed another outstanding quarter and continue to execute very well. We met or exceeded all key financial targets, delivered strong revenue and non-GAAP earnings growth, completed a $100 million share buyback and generated significant cash flow. Our first half was very strong and as a result, we-re raising our 2017 outlook for revenue, non-GAAP earnings and operating cash flow. Now to the numbers. As I talk through the results and targets, all comparisons will be year-over-year, unless I specify otherwise. Total revenue increased 12% to $680 million, at the high-end of our target range, with particular strength in hardware and IP. About 90% of revenue came from beginning of quarter backlog and 1 customer accounted for more than 10% of revenue. The weighted average license duration was approximately 2.7 years and we expect the 2017 average to be slightly less than 3 years. Total GAAP costs and expenses were $626 million which included a $38 million accrual related to the ongoing Eve-Mentor litigation. Total non-GAAP costs and expenses were $513 million, within our target range. Non-GAAP operating margin was 24.6% for the quarter. For the year, we expect solid organic margin expansion over 2016, moderated by the impact of the Cigital and Codiscope acquisitions. GAAP earnings per share were $0.34 and non-GAAP earnings per share were $0.88, at the high end of our target range. Operating cash flow was $123 million, driven by strong collections and business levels. We-re again raising our 2017 cash flow target to a range of $580 million to $600 million. We ended the quarter with cash, cash equivalents and short term investments of $1.1 billion with 11% onshore and total debt of $418 million. In Q2, we returned $100 million to shareholders through our stock buyback program…

[Operator Instructions]. Our first question will come from the line of Rich Valera of Needham & Company.

I wanted to ask about the strength you're seeing in the core EDA business. Aart, you've been on record on numerous occasions saying you thought that was kind of a low single-digit growth business, yet it looks like you're growing that close to double digits last couple of quarters. I just wanted to get your sense of what's driving that with any more color than you gave in your prepared remarks and kind of how sustainable this well above kind of industry growth rate might be.

Well, in general, I think we're in a phase of the industry around us and ourselves that tends to be strong and when we give you directions on growth rates of the different areas, we always look at it from a perspective of a multi-year outlook because that's sort of the only fair way to do it, all the more given that we have a multi-year agreement. Having said that, there's no question that the push towards more advanced chips is actually continuing strong and I think the reason for that is relatively straightforward which is that the new opportunities in technologies and products are using any of the existing computation and mobility techniques but now amended by increased digital intelligence requirements, will continue to be very strong after that enormous amounts of data generated by a happily growing set of sensors. And you can see that the entire field is really poised to demand more computation. And so it is natural, therefore, for people to come to our products. And while there are some up and down from quarter-to quarter, they are driving the leading edge and they need the tools to support that. So I think we're fortunate to do well.

Got it. And then just more specifically for the quarter, the service maintenance line was very strong. I'm not sure if you could give any commentary on what that was and how we should think about that line for the balance of the year.

Rich, on -- you're referring to the P&L with the split?

Correct.

On the services and product line, that includes the Cigital acquisition as well as strength in underlying IP, right? That's where the PoC revenues flow through as well.

But was that -- that was an unusually strong quarter. Should we think about that as a [indiscernible] ?

No, IP tends to be pretty lumpy, but we-re doing -- if you look at the results from a product mix, we-re doing well in IP. But it does tend to be lumpy and then you add that to the Cigital acquisition where we've got a full quarter of that now. That's driving a quarter-on quarter change.

Got it. So higher than historical levels but maybe not quite that high going forward. That makes sense.

Yes. We keep it in -- as we said publicly, the -- we expect IP to go grow in the double-digit rate and it could be lumpy from quarter-to quarter. But our goal is to drive that growth sustainably in the double digits.

Got it. And then just with respect to hardware you've obviously had a great run of demand here in hardware and it looks like Q3, probably not expecting any different, but you're currently modeling it, sounds like, for Q4 to maybe see a -- take a breather in hardware. Is that something you really have visibility to? Or is that just out of an abundance of caution? Just trying to get a sense of how much visibility you actually have to that Q4 sales hardware?

It feels like repetitive in the last 3 or 4 earnings releases. We must have used the word hardware is lumpy every time and that is an indication that the visibility is not complete, meaning that some of these things come in and then the customer wants immediate fulfillment or delayed fulfillment. So the timing can be somewhat arbitrary. There's no question that we have done well in the hardware area and I think that correlate directly to what we said earlier which is growth in complexity of chips but also more and more chips and the affiliated software needing to be verified together. And that is only a center of strength for Synopsys.

Rich, you bring up a good point in terms of Q4. The profile of Q4 is impacted by the variability of hardware. As you've seen in the results across geos, across product mixes, we're doing well. And going forward, you'll see more variability quarter-on quarter. So I think focus on the full year. And just as you brought up, the Q4 profile is down, but that's just mostly on hardware.

The next question is from Tom Diffely with D. A. Davidson.

First, I was curious what are the specific drivers of the hardware IP. Is it the core semi business, the system players or perhaps new customers in Asia?

Actually, the first two groups because in order for emulation and other hardware isolation methods to be useful, it is really in the proximity of the intersection of hardware and software which means it's the high-end chips that invariably drive a lot of software or it is the systems companies that integrate multiple chips around building apps or building entire product. And so the distinction between semiconductor vendors and system vendors sometimes is a little blurred because they tend to increasingly do stuff for each other or demand from each other that, that sits at this intersection. So both of those, I think, sort of grow hand in hand.

Okay. And then on the IP side, [indiscernible] the new customers are looking to fast track some designs now.

Yes. Actually, sorry I forgot that part of your question. IP is directly related to efficiency and productivity. And with continuous complexity of chips and by the way, continued complexity of the IP blocks, delayed collection of these blocks just continues to grow and the efficiency with which people can put together large chips is directly a function of which blocks are available, how easy is it to integrate them and how much do you trust the vendor because, obviously, you're highly relying on who provides you those blocks. And so as companies design large chips or larger systems, increasingly, they put their engineers on those aspects that they can't purchase but where they can provide additional differentiation while outsourcing increasingly blocks that, in the past, they would have done themselves. And I think that bodes very well for us.

Okay. So for both of those groups, too, have you begun to see any kind of a seasonality pattern with them? Or is purely just project based and can be fairly random through the year?

I have not seen any seasonality. I think it is very much project based and it has always a little bit the characteristic that the minute they figure out what's going to be in the project, they would like to have the IP the next day which, of course, not possible because a lot of this gets developed as part of agreements and service development sets. But the projects come and go. The companies have no aligned time line for these projects and so I don't see any seasonality. But one could have the configuration that, for some reason, there are some quarter, there are more project finishing than others or deliveries that we could do. So it could be somewhat lumpy, but it's not because of seasonality.

Okay. And then moving over to the core EDA market. Are the tools, the software tools now ready, completely ready for the 7- and 5-nanometer nodes? Or are there still challenges to be overcome?

They are completely ready for the 7-nanometer. 5-nanometer is not on the market and 5-nanometer is still in deep development. That will take a little while. And even when I say for the 7-nanometer -- 7-nanometer is really leading edge and what that means is that the foundries that provide the technology and that manufacture these chips will keep refining their technology on ongoing basis because they constantly are trying to tune the yields that they get because if you have a little higher yields, meaning a higher percentage of chips that work on a wafer, that has enormous economic impact. And so we-re in constant interaction with foundries. Every time that they tune it a little bit, we tune our tools and in some cases, we also tune our IP a little further.

Okay. And in your answer, they do include the IP and the 7-nanometers being ready as well?

Yes, we have a number of efforts in these advanced nodes and the word ready is not a black and white, meaning that there are certain things that are already earlier. There are certain things that are already with 7-nanometers, not quite there, for test chips, for example. And so there's a very complex multi-year process to bring something to market. But from my commentary, you should read that we-re well exercised in trying to align as much as possible with the foundry train so that we arrive at the station at the same time.

Okay. And does your answer there either include or not include EUV? Does it matter to you?

It doesn't matter that much. There are some very fundamental IP that's impacted by EUV. And as you may know, EUV is sort of now just slowly seeping into potentially some production manufacturing in the coming years and that is only for a few layers of the chip. And there are many, many layers, so EUV will be for the very refined ones initially. And so we're well on top of that, but it does not mean a big effort nor a big issue.

Okay. And then finally, Trac, you said that the accounting rule changes should have just not much impact on your model. If you were to see the impact, where would it be?

There are some portions of it in the IP business that could be affected and fortunately, we have got time to work through how to structure those contracts. And then there's a small portion on some elements of FSAs related to the software contracts, the FSAs being the equivalent to gift cards that you buy certain amount and then draw down over time. Those things will be challenging as we work through the rules, but fortunately, we know what they are. And they're relatively small portion and we've got some time to address them.

Our next question comes from the line of Krish Sankar with Bank of America Merrill Lynch.

I had a few more easy questions for you, Aart. Number one, on your IP business, can you guys say how much of it is actually ratable versus upfront?

Roughly 60% is ratable.

Yes. Just to clarify, when we do IP, people buy existing functionality but very often wanted to be delivered in a very specific form, maybe with some alterations or with some tuning for their versions of technology. And so this is why the IP is accompanied by a statement of work which can be viewed essentially as a service business. But it is service business around an existing product. And the reason you saw us hesitate is because we rarely think about it in those 2 categories. But it is a good question. It is sort of a balance between the 2.

Got it. That's really helpful. And then a couple of other ones. Obviously, there's been a lot of chatter about activity in China on their own semiconductor investment front. Have you guy started seeing any purchases for EDA tools from the indigenous Chinese companies?

Yes, we have seen those since like the middle 1990s. We've been in China for a long, long time and they have been, for now many, many years, many design companies. Many have not been successful. And so I would say, we're at least already on the third or fourth generation of companies where a bunch have disappeared, reconfigured and so on. And now there is a significant segment of strong Chinese companies that are -- that have, year after year, become closer and closer to be at the state of the art and we would certainly say that there are a few companies that are designing with absolutely state-of-the-art silicon technology. And so they have been broad users of our tools. China has been a growing market for us and we continue to see expansion of that market.

Krish [indiscernible] doing well across our entire portfolio. So we've been doing -- selling EDA in China for a long time as Aart described it. IP is a very good opportunity and we've seen growth there. And then even with the Software Integrity business, that's been emerging over the last few years as well.

Okay. All right. Let me -- I'm going to try to ask the question a different way then. If I look at, historically, for Synopsys, memory has been about 15% to 20% of your business and China has this huge plan to get into like memory with new 3D NAND factories. And from those customers, have you actually started seeing any -- are you guys having any conversation with this new 3D NAND investment in China? Or have you guys started seeing any early POs or anything of that nature?

Definitely yes. I mean, memory development is development that requires pretty deep understanding of semiconductor physics and device construction. And so with the number of companies we have now been working for a while on the development of those, you're absolutely right that, in the last recent period, I would say maybe last 2 years, there has been an increase in attention and effort. And that is the direct response to the fact that most economies are seeing an enormous amount of data coming about, that the big data will be analyzed and therefore, the -- be it in the cloud or be it the devices around it, are all going to use massively more memory and that their strong desire to make that memory a lot faster so that you can do more processing. And so in that sense, while China may be new to this game, the push in new memory technologies is quite interesting and that fits well the overall picture of a semiconductor business that is going to be an enabler to many other high-tech capabilities.

Got you. Got you. And then the final question I had is if I look at your fiscal first half or the last 2 quarters, your R&D as a percent of your sale, it's kind of been running at the low end of your historical range. I'm curious to know, is that the new fundamental shift on how you're going to be more effective with your R&D? Or are 2 quarters just not too big a data point to like extrapolate?

I wouldn't extrapolate off of 2 quarters. I think what you're seeing in first half is mostly the acceleration of revenues. We expect to maintain R&D investments in the 30%-ish range. And depending on how -- what you're seeing this year is just the lumpiness of revenues and the quarterly profile will skew that. But overall, we expect to maintain the model in that 30% range.

And our next question will come from the line of Sterling Auty, JPMorgan.

I'm wondering, with the acquisition of Cigital on top of all the other elements that you've pieced together in some of the software assurance, et cetera, how much of your sales and marketing headcount is now focused outside the core kind of systems and semiconductor industries.

Good question. I don't know the exact answer, but I can venture that it is probably half to a little bit more as outside. And the reason for that is that the marketing needed in our core semiconductor space is not that high because we have many very good connections and we can relatively easily access people that need security and quality tools for their software. Whereas in the rest of the space, there are many different market segments. And you mentioned Cigital which is a really interesting acquisition because it's not only that it brings to us a service business, that allows us to engage at a higher level with customers that are looking at the strategy of their -- of how to deal with quality and security. It is that they're also very strong in the financial market. And so the way you communicate with financial market is very different than, let's say, the health market or the embedded software market and that is why some marketing effort has to be specialized on that. From a sales point of view, there, it is much more a question of global access and how quickly can we grow this and be in a situation where we have high-quality salespeople. And that is an ongoing effort and going well. We're growing well, but there's no end to that one yet.

And then on the IP business, you mentioned the strength that you're seeing there. We look at the gross margins. Obviously, the gross margins relative to expectations was probably heavily influenced on the mix. But specifically to the IP, how would you characterize the gross margin contribution at this point and the thoughts of where that could trend?

Sure. This is Trac. The margins on that are pretty healthy, but directionally, they're slightly below the corporate average. And that's about where we'd expect to drive that.

Okay. And then last question, you mentioned 606 on a couple of areas, but I didn't hear one of the attributes of 606 is the backlog disclosure. Any thoughts around the structure of how you're going to disclose the backlog? And specifically, I know it has to be done quarterly, but there is the commentary that goes with it describing how you go from backlog into revenue. Is there any conclusions at this point in terms of how that would be -- how those disclosures will come?

No. We're still assessing that, but I would say that we've done a pretty good job disclosing backlog in the past. And the metrics that we provided, I think, gives investors a pretty good sense of the health of the business. Going into any particular year, we do aim for about 80% visibility of backlog, scheduled revenue in backlog at the end of the year. Secondly, we try to enter the quarter -- each quarter with about 90% of backlog in hand. So I think those 2 metrics combined with anything else we provide should give pretty good color on the health of the business.

Next question comes from the line of Jay Vleeschhouwer with Griffin Securities.

Trac, let me start with you. You raised the midpoint of your cash flow expectation for the year by $80 million. Is that pretty broadly based in terms of numbers of customers driving that? Or is it fairly narrowly based in terms of perhaps there's a small number of customers with a large amount of collections driving that number? And then secondly, for you, before I turn to Aart, given the scaling that you're seeing now in hardware, could you comment on the margin or mix trends vis-à-vis hardware profitability?

Okay. Let me start first with the broad -- the cash flow, whether it's broad based. It was, yes, definitely broad based and that would be consistent with the health of the business that we're seeing in the first half. As I mentioned, when you look across, the growth was pretty broad based across geographies and product mixes. So that -- the cash flow is consistent with that. And then from a hardware perspective, we-re definitely doing well with hardware revenues and the profitability on that certainly is improving and it is [indiscernible] profits.

Is it improving as a percentage of the hardware revenue?

Yes.

Okay. For Aart, a couple of things for you. We often hear over the years from EDA companies when new hardware is introduced, particularly emulation of course, detailed descriptions about the hardware specifications and most particularly about the design capacity that the hardware systems are capable of. We don't typically hear that sort of discussion around the software tool capacity. And this issue has come up for you and others over the years in terms of design size capacity that you could put through, DC, for example and the issues of that kind. As you move now to 10-nanometer and below, could you talk about whether or not there are any capacity issues that you need to resolve for any role of your tools to adequately handle the design sizes that you're likely to see? You mentioned new version of VCS, so perhaps that's one example. But anything you can address with regard to the capacity on the software side would be helpful. And then lastly for you on the technology front, are you beginning to see any kind of conjoined usage of your prototyping hardware with software integrity, in other words, is the software to bring out capability of hardware being used somehow with the Software Integrity business? And then similarly, is the prototyping business being used increasingly along with emulation as part of your Verification Continuum?

Okay. Well, so for being at Synopsys for 30 years, I would say our capacity has always been less than the customer wanted and more than they thought they would get, meaning that the state of the art is limited typically by a few things which is, one, is how fast can you run; and two, how much can you run at the same time. And so it has been actually quite remarkable how, over all this time period, these numbers have continued to increase. And I'm certainly on record of having argued many times that we-re the other half of Moore's Law which is if they can manufacture it, we will find a way to simulate it, to design it and so on and that goes hand in hand. But for the most state-of-the-art situations, one has always had the limit of what one can do. The second comment I would have for that, it is quite remarkable how, over all these years, over and over, there have been redesigns of architecture of utilization of compute environments, of utilization of multiple computers, multiple cores, multiple threads that have managed to keep your question in check. So we're always chasing the horizon in other words and this is true on the hardware side in emulation or prototyping. It is certainly true for software for a long time. And expect that to absolutely continue and we will stay at the leading edge of that. Secondly, regarding the intersection of prototyping and software integrity, I expect that we're going to start seeing some of that in the not-too-distant future. So far, there's not been a lot of evidence of that, but we have not really pushed on that front yet either. The intersection between prototyping of emulation, we have already seen quite a bit of that. And by the way, it's intersection of prototyping, emulation and simulation and in some cases, even virtual prototyping. And that is because the system people would like to simulate software on larger systems that have many components that may be defined at different levels of abstraction. So all of this is being done. I don't want to give the impression that this is routinely easy. This is sophisticated stuff. But we-re in a very strong position with this and the most advanced customers are very literate about this type of problems.

Okay and our next question will come from the line of Monika Garg with Pacific Crest.

First, just if I look at your guidance, even if you've modeled the high end of revenue guidance, since Q4, you were guiding almost $50 million down Q-over-Q. Historically, your Q4 is at least flattish to modestly up. So what is leading to such high kind of revenue down Q-over-Q? Or is it just you're being conservative?

No, Monika, it's really the hardware business. As we mentioned, it will vary from quarter-to quarter. The underlying -- outside of hardware, the business -- all the other businesses are doing well, so it's not offsetting anything. But hardware is driving the quarter-on quarter comparison.

$50 million?

I'm sorry?

So the downside is about $50 million, 5-0. The others are up.

We have very strong hardware business.

Okay. And a similar point, your operating margins, if I have to model for the guidance, it's going to be like 17%, again, down like almost 700, 800 basis points year-over-year and Q-over-Q. Hardware is down. Operating margin should be better, right?

As we mentioned at the earlier comments, expenses are seasonally up in Q4, so we've got the offset of revenues coming down because of the hardware trend and the expenses going up in Q4.

All right. Then, can you just remind us how big your software security business now? And when do you expect it to be breakeven?

Well, as we mentioned last -- in previous calls, we were guiding to about $100 million last year which we met and a -- the overall growth in that business is about 20% which is tracking, too. And then Cigital, when we acquired them, was roughly half of the existing business. And then the underlying business, we mentioned that the acquisitions would be dilutive to earnings this year and would affect margins. But on the classic software integrity business, we-re on track to do breakeven at second half of the year.

All right. Then, going forward, what is the growth rate you expect in the software security business, including the Cigital acquisition?

We're looking at the 20% range.

Okay. Then, Aart, we saw like, over the weekend, a big security breach. Could you maybe talk about how this could impact the security -- software security business? Have you seen any...

Yes. For starters, we're fortunate that our own processes around upgrading our firewalls, of making sure that all of the patches have been applied to software have been effective, so we have not had any issues with this. Very clearly, every time there is a big story, in this case, massive number of stories, the attention level around the notion of security goes up. And we-re sitting in the space where the idea is to design security in the first place, not to just fix all the things that went wrong in the past but make it essentially correct by construction. And so from that perspective, I think we-re in a very healthy situation. If you amend that by the fact that Gartner which tracks many, many security-related companies such as ourselves, are now putting us in the leading quadrant for the activities that we do, obviously, the attention that this is getting is positive for our business even if it's negative for the world and so we-re seeing that with the acquisition of Cigital, with the commentary that Trac just gave you on the existing business. I feel that we can now state that this business has, for lack of better term, reached critical mass. And what I mean with that is that a number of years ago, we made the decision to move up from the silicon to the software. We made a decision to bank beyond the hardware and software intersection and that, moreover, this new TAM that touches many companies that, in the past, we have no business with looks very promising. There were many execution challenges in getting there. I think we've navigated well through those. And I'm very encouraged with where we-re in the middle of 2017 because we see this vision materializing. And hopefully, it's not on the back of catastrophes around us but more on the premise that building correctly for the future is really where the software industry should go and that is where our contribution will be.

Got it. Then, the new administration has talked about making it easy for cash repatriation. If that happens, how could you think about the usage of cash?

Monika, we-re staying on top on what's developing as far as tax reform in the U.S. If it does occur, I mean, obviously, we -- I think when you use cash the same way we've been using over the last several years. Over the last, trailing 12 -- 4 quarters, we've bought back close to $400 million in stock. We've acquired companies in the software integrity space and we'll continue to use the cash to drive growth in terms of acquisitions and then return cash when it's appropriate via buybacks.

Last one for me on IP, growing double digit. Is double digit the right way to look at going forward? And then you've been building ADAS IP for autos for some time. Are you seeing inflection in that?

Well, for the first part of your question, yes, we think that it is in the low double digits. So far, the team has been executing particularly well and if you look at it from the complex chips, our expectations are that we will continue to see good growth and healthy business there. I'm sorry, I couldn't acoustically hear your second part of the question.

You've been building the ADAS IP for autos for some time. Maybe are you seeing inflection in that?

Oh, yes, sure, sorry. Yes, automotive has been an interesting area because it's having so much attention, frankly, by a whole number of people that may create a bit of a bubble there. But at the same time, there's a good reason it's getting so much attention which is it is the showcase for what modern hardware software will be able to do to fundamentally transform a field that has been relatively slow in evolving. It is now evolving very, very rapidly. And so we have invested quite a bit over the last few years not only to make sure that the tools are capable of dealing with this type of designs, but our IP portfolio has grown significantly in the realm of having certified IP. And certification of IP is for one key reason which is automotive demands a certain set of checks to make sure that what you're doing is actually safe. And in the past, safety has been actually developed quite well in automotive. With this very, very rapid increase of both hardware and software, the complexity that comes with that is going to be a challenge for safety and the fact that we have a portfolio that is now massively certified, I think, bodes very well for our position there.

And our last question comes from the line of Mitch Steves with RBC Capital Markets.

Two questions for you if I could. First for Aart. Just in terms of the core EDA business, that's continuing to grow at double digits now. So I know you guys have talked to low to mid-single growth. But what exactly has changed there? I mean, I know the CAD business or the CAD industry to be growing faster. Could you maybe highlight some changes that have happened over the past year or so?

Well, one of the key pieces is that the intersection of hardware and software has been driving verification extremely hard and a subcategory to that is really emulation and prototyping. We refer to that as hardware. And that growth has just been exemplary because so many people would like to run software on hardware they don't have and so instead, they use an emulator to sort of fake the hardware if you like. And I expect that to continue because if you want to go to market, you want to start the software development as soon as you possibly can and make sure that it's going to work with the hardware that takes a long time to manufacture. So that has really been one of the key variables in this game.

Got it. And then secondly for Trac on the financial side. I'm having a hard time getting to the kind of Q4 EPS because if I look historically, typically from Q3 to Q4, your total expenses go up by $20 million or so but in order to get to the new midpoint that the full year guide implies, if you're going to be $30 million plus or a 50% increase, so I guess, what's kind of in the financial model for Q4 that's abnormal from the past?

Well, normally, the -- as you highlighted, the expenses do ramp up from Q3 to Q4. What we would normally see this year is an increase in hiring. That's per our plan. You would see a ramp-up in variable comp as we accrue for where we think we're going to end up the year. And then, historically, the numbers I think -- well, we can model with you, offline, but your historical numbers in terms of growth might be a little aggressive. And then the offset to that is how you model the Q3 revenue to Q4 profile, but we can work with you afterwards.

Okay. And we have no further questions in queue and we do have 5 minutes remaining. I'll turn it back to the executive board.

Well, thank you very much for attending this earnings release. Hopefully, you took away that we had a very strong quarter, that the outlook for the coming quarter is strong, the overall year is looking extremely solid. But most importantly, I think the alignment around our Silicon to Software strategy appears to be working quite well and all the individual businesses have contributed strongly to our present success. So with the intent to continue on that trajectory, we thank you for your support and attention.

Ladies and gentlemen, that does conclude today's conference. Thank you for your participation and for using AT&T. You may now disconnect.