Good afternoon. My name is Erica, and I will be your conference operator today. At this time, I would like to welcome everyone to the Cadence third quarter 2018 earnings conference call. All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question and answer session. If you would like to ask a question during this time, simply press star, then 1 on your telephone keypad. Thank you. I would now like to turn the call over to Alan Lindstrom, Senior Group Director of Investor Relations for Cadence. Please go ahead.
Thank you, Erica. I would like to welcome everyone to our third quarter 2018 earnings conference call. I am joined today by Lip-Bu Tan, CEO, and John Wall, Senior Vice President and CFO. The webcast of this call is available through our website, cadence.com, and will be archived through December 14th, 2018. A copy of today's prepared remarks will also be available on our website at the conclusion of today's call. Please note that today's discussion will contain forward-looking statements and that actual results may differ materially from those expectations. For information on the factors that could cause a difference in our results, please refer to our filings with the Securities and Exchange Commission. These include Cadence's most recent reports on Form 10-K and Form 10-Q, including the company's future filings and the cautionary comments regarding forward-looking statements in the earnings press release we issued today.
In addition to financial results prepared in accordance with generally accepted accounting principles or GAAP, we will also present certain non-GAAP financial measures today. Cadence management believes that in addition to using GAAP results in evaluating our business, it can also be useful to review results using certain non-GAAP financial measures. Investors and potential investors are encouraged to review the reconciliation of non-GAAP financial measures with their most direct comparable GAAP financial results. The reconciliations are available at the investor relations section of cadence.com. Copies of today's press release, dated October 22nd, 2018, for the quarter ended September 29th, 2018, related financial tables, and the CFO commentary are also available on our website. Now I'll turn the call over to Lip-Bu.
Good afternoon, everyone. Thank you for joining us today. I'm very pleased to report that broad-based customer demand across our core EDA, hardware, and IP product lines enable Cadence to achieve excellent operating results and financial performance in the third quarter. We live in a data-driven world that is propelled by key technology waves such as cloud data center, 5G, and machine learning. AI machine learning, fueled by big data and sophisticated data analytics algorithm, along with the move to domain-specific accelerators, is transforming many industries, including transportation, healthcare, and manufacturing. All of this lead to increasing demand for high-performance compute, high-speed connectivity, and dense storage, which in turn drives strong design activity and broad demand of our innovative system design enablement solutions. Our SDE strategy continues to open up new growth opportunities as we expand our focus beyond semiconductors to systems and customer in newer verticals.
As a result of these factors, we are increasing our outlook for the year. John will say more on this in a moment. I will now review the Q3 highlights, starting with IP business. Our IP business had a strong quarter with double-digit growth, driven by increasing royalties for Tensilica and robust demand for our memory products. Additionally, we launched two important new IP products. First, earlier today we announced the industry-first silicon-proven 112G Long-Reach SerDes IP in seven nanometer technology. This innovative SerDes technology based on nusem i acquisition is essential to enable next-generation cloud connectivity to move to 200 gigabits and beyond in the hyperscale data center. We have been working closely with early adopting customers and are already taking orders as we begin to engage broadly with customers.
Along with our leading DDR and PCIe solutions, we now offer the most compelling portfolio of essential IPs for the hyperscale data center. Second, we also significantly enhance the Tensilica product line with the DNA 100 processor, which is a deep neural network accelerator. Its innovative architecture enables the DNA 100 to deliver leading performance and power efficiency while scaling across a broad range of compute needs. The DNA 100 is ideal for embedded inferencing applications, where latency and lack of connectivity may be an issue, such as surveillance, drones, AR, VR, and automotive sensor fusion. Next, I will discuss highlights of our system design and verification solutions. The Cadence Verification Suite had another strong quarter, with year-over-year revenue growth of 9%. Palladium Z1 also had a strong quarter as demand for increasing hardware capacity continues.
Two customers significantly expanded their installation of Palladium Z1, and overall, we added five new logos. Our new Palladium Cloud offering, which provides on-demand, cloud-based emulation capacity, is ramping up nicely and contributed meaningfully to orders in Q3. Adoption of our Xcelium simulator continued to grow as we added several new customers, including a large commitment to our technology from a market-shipping customer. Digital and sign-off continues to perform well, with revenue up 9% year-over-year. Proliferation of our digital and sign-off solutions continue with existing market-shipping customers, as well as wins with the new customers. We added 19 new logos in Q3 for our digital and sign-off products. Customers tape out more than a dozen seven-nanometer designs in Q3 using Innovus. In total, more than 50 customers are using Innovus for implementations at the seven-nanometer node.
We are actively engaged with very early adopter customers on their five-nanometer designs and have started work with partners to get ready for three nanometer. We won four Partners of the Year awards at the TSMC Open Innovation Platform, including on five-nanometer design infrastructure collaborations. Last quarter, we introduced Cadence Cloud in collaboration with major cloud industry players: Amazon Web Services, Google Cloud, and Microsoft Azure. Our offerings have already been deployed in production with customers. Interest in using cloud for semiconductor design is growing. We are pleased with the customer receptions and have a healthy pipeline of opportunities. In October, Cadence collaborated with TSMC to launch their virtual design environment. Cadence Cloud was certified as a storefront for mutual customer desiring to develop SoC using TSMC IP in the cloud.
The endorsement is further evidence of the growing interest in using the cloud for semiconductor and electronic system design. Before turning it over to John, let me quickly summarize comments. Cadence achieved excellent results through consistent execution of our system design enablement strategy across our core EDA, IP, and hardware businesses. IP had an excellent quarter, and we announced two exciting new IP products. Important market-shipping customers expanded their use of both the software and hardware-based products in our Cadence Verification Suite. Our digital sign-off solution continue to proliferate with market-shipping customers, and we are actively engaged with customer and partners on five nanometer and three nanometer development. I will now turn the call over to John to review the financial results and provide our updated outlook.
Thanks, Lip-Bu, and good afternoon, everyone. As Lip-Bu said, broad-based customer demand across our major product lines enabled Cadence to achieve excellent operating results and financial performance in the third quarter. I am pleased to report that we exceeded all of our key performance metrics. We expect strong demand and cash flow to continue into the fourth quarter, and as a result, we are raising our outlook for fiscal 2018 and increasing stock repurchases to $75 million for the fourth quarter. Before we get into Q3 results, I'd like to remind you that Cadence adopted a new revenue accounting standard known as ASC Topic 606 for fiscal 2018. These new rules, as we often refer to them, are now GAAP for Cadence. The numbers I present for our third quarter are based on these new rules, unless otherwise stated.
Please also keep in mind that this is our transition year to the new rules, and our results under the new rules are not directly comparable to those of 2017, which were reported under ASC Topic 605, or the old rules. To provide a more direct comparison against our 2017 results, we will show our quarterly results as reported under the old rules for all four quarters of 2018. Let's go through the key results for the third quarter, starting with the P&L. As reported, total revenue was $532 million. Non-GAAP operating margin was 32%, GAAP EPS was $0.35, and non-GAAP EPS was $0.49. Under the old rules for direct comparison against Q3 2017, total revenue was $526 million, non-GAAP operating margin was 32%, GAAP EPS was $0.34, and non-GAAP EPS was $0.49. Let us turn to the balance sheet and cash flow.
Cash and short-term investments totaled $550 million at the end of Q3, with approximately $135 million of that cash here in the U.S. Debt outstanding at quarter end was $350 million. Operating cash flow in Q3 was $110 million. During the quarter, we repurchased $50 million of Cadence shares and paid off our existing $300 million term loan. As reported, DSOs were 42 days. Under the old rules, DSOs were 39 days. Looking ahead, we expect strong demand to continue into the fourth quarter. As a result, we are increasing our outlook for fiscal 2018. For Q4, we expect the following results. Revenue in the range of $545 million-$555 million, non-GAAP operating margin of 29%-30%, GAAP EPS in the range of $0.27-$0.29, and non-GAAP EPS in the range of $0.46-$0.48.
For fiscal 2018, we now expect revenue in the range of $2.113 billion-$2.123 billion, non-GAAP operating margin of 29.5%-30%, GAAP EPS in the range of $1.15-$1.17, non-GAAP EPS in the range of $1.80-$1.82, and operating cash flow in the range of $550 million-$580 million. Our Q3 results and outlook for the second half are significantly better than we expected this time last quarter. Design activity is healthy, and we experience broad-based customer demand across our core EDA, hardware, and IP product lines. Upside in our hardware and IP product lines are the primary contributors to the upside in our revenue guidance for 2018. Some of our customers requested earlier delivery of hardware systems, and we saw an uptick in our IP royalty revenue.
Our second half also has upside resulting from a number of one-time benefits on the expense side, almost all of which fell into our third quarter. The upside in our expenses for Q3 was primarily due to timing of new hires and one-time credits to professional services expense. For fiscal 2018, we now expect the difference in revenue under the new and old rules to be approximately $12 million, most of which is due to IP. This means that under the old rules, our implied 2018 guidance at the midpoint is now expected to be revenue of approximately $2.13 billion, non-GAAP operating margin of approximately 30%, GAAP EPS of approximately $1.18, non-GAAP EPS of approximately $1.85, and operating cash flow in the range of $550 million-$580 million. We expect our operating cash flow to be the same under both the new and old rules.
One thing I'd like to point out is that our fourth quarter earnings release will be scheduled for February 19th, 2019, due to the additional work required to complete our year-end accounting for the first time under both the new tax law and the new revenue accounting rules. To sum up, I am very pleased with our progress in 2018. It is shaping up to be a great year for Cadence, and I'd like to thank the entire Cadence team. Their operational discipline and their drive and passion to make our customers successful are truly inspiring and have played a large part in allowing us to raise our guidance throughout the year. On an apples-to-apples basis, we are now expecting annual revenue to increase by more than 9%.
On the back of that revenue growth, we now expect non-GAAP operating margin on the basis of the old rules to improve to approximately 30% for the year. With that, operator, we'll now take questions.
At this time, I would like to remind everyone, in order to ask a question, please press star followed by the number 1 on your telephone keypad now, and we'll pause for just a moment to compile the Q&A roster. Your first question comes from Gary Mobley from Benchmark.
Let me first extend my congratulations to Lip-Bu for your 10-year anniversary.
Oh, thank you.
I want to start out by asking a question about Cadence Cloud. Can you give us a sense of what % of your fourth quarter revenue outlook might be generated from Cadence Cloud? Then as well, maybe kind of a preview into fiscal year 2019. John, last time we had this call and I asked the question, you mentioned you don't expect a material difference in revenue recognition with Cadence Cloud versus traditional customer delivery. Is that still the case?
Yep. I think you have two questions. Let me answer the first one. First of all, last quarter we introduced the Cadence Cloud with collaboration with Amazon, Google Cloud, and Microsoft Azure. We are very delighted that right now our offering are already deployed into production by customer. Interest using the cloud for design is growing. We are very excited about the customer reception, and we have a healthy pipeline opportunity. I just have to say that we are in the very beginning of the ending for the baseball, right now in the World Series now. We are still in the beginning of the ending. The early October, we are delighted to collaborating with TSMC launch their virtual design environment.
As you know, moving to the cloud foundry partners is important, so they make sure that their IP are also in the cloud, the TSMC Cloud offering, so that customer can use that and using our tool in the cloud to developing the SOC. This is very exciting. Overall, we excited about the opportunity, but in the very beginning of the ending.
Gary, just to address your second question, there's no material impact from the Cadence Cloud on Cadence revenue. Although Cadence Cloud is a valuable addition to our product portfolio for both customers and Cadence. It's not a material impact to our revenue in Q3 or expected to be a material impact in Q4.
Okay. As a follow-up question, I wanted to ask about the memory IC industry. I think it's widely known that we've seen some pretty steep drops in NAND flash pricing and maybe a stagnant DRAM market. I know historically memory has not been a heavy user of EDA tools, but can you give us a sense of what % of your revenue comes from memory IC companies?
I don't think we disclosed the memory IC percentage of our revenue, I think clearly, in this data-driven world, it's all about big data and then also how to do the data analytics beside the compute, because the workload have changed. I think, this memory and in-circuit memory, some of this application become very critical. We have a very strong footprint with the memory IC player, and we are delighted to work closely with them for their next generation's design. Some of this, as you know, the NVM, non-volatile memory side, is taking off in the hyperscale side. It's all about data. I mean, from the IoT to the edge, to the automotive driving, the ADAS, and all the different IoT for the Industry 4.0, data become very essential.
How to address the latency and the speed of time to get the data and then able to make some intelligent decision, those are critical. I think that's a lot of disruptive innovation R&D is working on, and we are very well-positioned with the memory key players, and we're delighted on the tool and the IP front to work closely and close collaborating with them.
That's it for me. Thanks, guys.
Thank you.
Thanks.
Our next question comes from Mitch Steves from RBC Capital Markets.
Hey, guys. Great quarter, and thanks for taking my question. I had two. First one is actually on the ASC Topic 606. I remember at the beginning of the year, you expected a $40 million impact. Now it sounds like it's a $12 million difference, and it seems like the total impact is almost positive for you guys. Maybe you can help us understand what happened there and what changed throughout the year.
Yeah. Mitch, this is John. The difference is down to $12 million, and it's mainly due to how IP revenue is recognized under ASC Topic 605 and ASC Topic 606. Under ASC Topic 606, the new rules, we record IP revenue as we deliver the IP to the customer. Under ASC Topic 605, the old rules, some IP cannot be recognized until all of the IP committed in the contract is delivered to the customer. As a result, we recorded IP revenue during Q2 and Q3 under the new rules that will show up in Q4 revenue under the old rules. Just for clarity, we expect the revenue under the old rules to be $12 million higher than revenue under the new rules. Just want to make that clear.
Okay. Got it. Thank you. Then the second one is actually on the overall macro, and we've seen some negative news about semiconductor volumes. I wanted to see if anything has changed in terms of the R&D you guys are seeing, or in terms of the engineering hires you're seeing in the space, just broadly from a macro perspective.
Sure. Mitch, let me answer the questions on the environment. Clearly, the design activity is quite healthy and basically driven by a couple of things that I mentioned earlier: the data-driven and with the big data and the machine learning, deep learning, that using the big data and then do the data analytics; and also kind of moving towards this, I call it the domain-specific application that's very broad application to the very multiple industry, from transportation to healthcare drug discovery, manufacturing, and of course, the automotive-related area. I think this all driving more and stronger design activity from the big companies and also the small company from the service provider. They want to optimize and differentiate their service, so they are starting to quietly building up silicon development. We are embracing and partnering, collaborating with them deeply.
Of course, our SDE strategy is starting to really play big time in term of addressing some of this new growth opportunity, so that we can really focus on helping the customer to design that. I think overall, answer your question, all these are driving the design activity. We see a very nice increase on our collaboration with our customers.
Perfect. Thank you.
Thank you.
Thanks.
Our next question comes from Monika Garg from KeyBanc.
Hi. Thanks for taking my question. First, if you look, you're guiding almost 10% growth on ASC 605 basis. How to think about growth next year?
First of all, I think, Monika, as you know, that we don't provide guidance for next year. Wait for January, and we will provide that. John?
Monika, I wouldn't focus too intensely on any one quarter. I mean, revenue growth should be strong in Q4, but in Q3, we had a number of one-time benefits that don't follow through into Q4. Expenses in Q3 benefited from the timing of new hires and some one-time credits to professional services expense. That said, I'm very pleased with what we are expecting in non-GAAP operating margin of approximately 29.5%-30% for this year.
All right. Lip-Bu, could you maybe talk about how do you see impact of tariff impacting semi industry? Could you see EDA industry, any impact from that?
Can you repeat your question again? I missed the first part.
We are seeing a kind of tariff between U.S., China, and other geographies. Your comments regarding how could that impact semi industry, in turn, could you see any impact on the EDA industry from that?
Yeah. I think it's a very important topic, that the tariff landscape is very fluid and very difficult to predict. With our guidance is everything we know also reasonably estimate. Clearly, we watch that very carefully. Overall, we have done well in China, then China is a very growing opportunity for us. We have a big team over there to support our customer. At the end of the day, I think we are just really laser-focused on supporting our customer globally and in their design activity.
All right. There's just the last one. John, operating margin's very strong at 30%, close to 30%. Where do you think operating margins could be next 2-4 years? Thank you.
Oh, nice try, Monika. Yeah, we're not guiding beyond this year for 2018.
Thank you so much.
Thank you.
Our next question comes from Richard Valera from Needham & Company.
Thank you. Let me add my congratulations for the strong results, gentlemen. I think this is a question for John around, you mentioned you had some one-time benefits and expenses in Q3, and also, I think you mentioned early hardware shipments, which may have contributed to the strength in Q3. I wanted to just say, if you looked at the year as a whole, including Q3 and Q4, is there anything we should think of as one time that wouldn't make for a clean comparison to 2019, even though I know you're not guiding, just is there anything we should sort of back out as we're looking at 2019?
Yeah. Good question, Rich. What I would highlight is that we delivered hardware in Q3, for which we won't collect payment until the following quarter. If you have a look at our CFO commentary, we're expecting that to happen again in Q4. In our CFO commentary, we've called out that we expect DSOs to rise to about 45 days at year-end. That's mainly because some customers have requested delivery of hardware in 2018, for which we don't expect to get paid until early 2019. Yeah, taking the year as a whole, that's what I'd be inclined to point out.
If you looked at it, coming into this year versus last year, because you actually had a really strong, I think, bookings hardware quarter in the fourth quarter of last year. Did you have that same effect where you had kind of deferred hardware revenue coming into the year that was recognized early this year, and now you kind of have that same impact likely going into 2019? Is that fair?
What we saw at the end of last year was an acceleration of hardware bookings into the end of the quarter, at the end of 2017.
Right.
And like I say, where that contrasts with 2018 is, this year some customers have requested earlier delivery of hardware in Q4 2018, for which we're not expecting to get paid until Q1 2019. With all that said, we're not guiding beyond the end of 2018.
Okay. You will see some revenue in presumably the first quarter of 2019 from hardware that you'll ship in the fourth quarter of 2018. Is that fair?
We trigger revenue on delivery of the hardware. If we ship hardware in Q4, we'll recognize revenue on that hardware in Q4.
Okay. Thank you for the clarification. Appreciate it. Asia Pac seemed to show up as a notably strong geography in this quarter. Is there anything there that we should be aware of, or is that just noise in the numbers?
Yeah. We continue to see strong growth in Asia. For the quarter, the growth, if you look across the business groups, it was mainly in most of the revenue upside for the second half is in verification and IP. That's mainly because those segments benefit from upfront revenue recognition on delivery of the IP and hardware. We had strength across all product lines.
Got it. Okay. Thanks for taking my questions, gentlemen. Appreciate it.
Okay. Thanks.
Our next question comes from John Pitzer from Credit Suisse.
Yeah. Good afternoon, guys. Thanks for letting me ask the questions. Congratulations on the strong results. John, I apologize if I missed this, but relative to the one-time benefits of OpEx in Q3, did you quantify them across SG&A and R&D? I'm just trying to figure out relative to your initial guidance of op margins being down over a couple of hundred basis points sequentially, they were up, and I'm just trying to figure out what drove the difference and how I should think about sequential op margin growth into the fourth quarter.
Yeah. Fair question, John. What can I say that the second half is just really strong for us. There's been no change in how we do our guidance or anything. It's just design activity is healthy. I wouldn't read too much into any one quarter or even one half. A lot of the things that are going in the right direction for this year, especially in the second half, on both the revenue and expense side. We're operating the business for the long term. Q3 in the second half does include, as you said, a number of one-time items. Even adjusting for those, I'm very pleased with how the second half of the year is playing out.
John, is it fair to say that actually adjustments that maybe operating margins would have been about half as strong relative to guidance? If they were 500 basis points above, about half was one-time, half was operational? Can you comment out there?
Yes, of course. Sorry, John. Yes, that's correct. About half was due to the one-time expense items and half due to the revenue growth.
Perfect. That's helpful. Maybe as a follow-up, Lip-Bu, just going back to Cadence Cloud. If you think about the long-term potential for this distribution channel, how big do you think it becomes as a % of revenue? Is this something that you think your traditional customers exploit, or is this really an avenue to kind of grow the customer base of design activity? You talked about rev rec not being changed by Cadence Cloud. Is the dollar of opportunity different between buy versus cloud when you think about project-based revenue?
Let me answer the first question, and then John can answer the second question. On the first question, as I mentioned earlier, this is kind of an early ending, but we are very excited with the customer receptions and also with a healthy pipeline opportunity. Clearly our partner with industry leader in the hyperscale cloud have been really very good, and we are very happy with their collaborations. Clearly, if you look at from the design point of view, if you can partition and then over the multiple unlimited server that cloud infrastructure offer, clearly the PPA runtime improvement is substantial. We already see that in benefit on that, and the customers see that. In some way, we can make the faster and better performance for customer. I think stay tuned.
We're going to continue to drive that, especially with the TSMC virtual design platform and using our tool and also their IP in the cloud. That can be very exciting, and we're going to do that also for other foundry partners so they make it available to our customer, whichever way they pick on the hyperscale, whichever they pick on the foundry, we will be there to support them, and our tool will be optimized for their solution design.
John, to take the second part of your question, Cadence Cloud doesn't change our business model. It offers customers another way to optimize their investment in Cadence tools. In saying that, the Palladium Cloud is probably the best opportunity in the nearer term for incremental revenue because it taps into customers who traditionally have not had the capital budgets to purchase emulation hardware. We had some revenue contribution from Palladium Cloud in Q3.
Guys, if I could sneak one more quick one in. Lip-Bu, what % of your revenue today is domestic Chinese? I guess just relative to trade war concerns, do you see any evidence of those customers perhaps ordering more than they need for fear that the trade war escalates?
Yeah. I don't think we break down on the % of the customer from China domestic customer. Clearly, we are engaging quite heavily. Our philosophy is to support the leading customer in their most complicated design globally. China is included, and there are some world-class company from China. We've been collaborating closely with them. They like our true partnership trusted arrangement, with the best design on the tool and IP. We're going to continue doing that. So far have been okay.
Thank you.
Thank you.
Our next question comes from Sterling Auty from J.P. Morgan.
Yeah, thanks. Hi, guys. I think the strength in the upside in the quarter, as you mentioned, was both verification, emulation, as well as IP. On the verification emulation side, can you give us a little bit more color? Is this existing customers buying more? Is this new customers? What was the balance? Just what's the source of the strength that you saw in the quarter?
Yeah, let me start it first, Sterling. A couple of things. One, clearly the customer increasing the demand on the hardware capacity continue. We mentioned two large customer significantly expand their installation of Z1 purely in a multiple factor, but one of the key reason is a lot of complex design. They need more capacity for verifications. Talking about that, clearly also driving our Cadence Verification Suite, we're delighted on the Xcelium side, we have one large commitment to a technology from a market-shipping customer. I think all in all, I think it's a good quarter across the hardware emulation Z1 and also Xcelium. We continue to drive large scale design. This is a must-have, and they're able to scale.
Clearly, it's a good balance between the existing customer and also our new logos that we highlight, that we have five new logos. The customer that are new customer to us, they see the benefit of the usage of our Z1.
When looking at the fourth quarter, I want to make sure I'm clear. The guidance that you gave, is the driver for the new guidance, again, more kind of the emulation IP, so things that have upfront revenue recognition that's driving the fourth quarter changing guide, or am I missing that?
Sterling, this is John. Yeah, revenue exceeded internal expectations across all our major product lines. Yes, most of the revenue upside is in verification and IP when you look at the half, and that's because those segments benefit from upfront revenue, as you just said.
Sterling, just to add on, I think we mentioned it's a broad base, EDA, hardware, and IP. By the way, the digital and signoff have a wonderful quarter, 9% growth. We have mentioned about more than 12, the 7 nanometer design win on the Q3. Clearly, we go deep into the 5 nanometer with our partners and customer. We are getting ready on the 3 nanometer. Continue driving the improvement on the various different tool. It's a very broad base.
Okay, great. Last question, just to clarify. The previous caller, John, you mentioned Palladium Cloud. I thought there would be a difference in revenue recognition going from upfront when you ship a box under current model versus I thought the cloud model would be more of a ratable or transaction based for Palladium Cloud.
Yeah, that's a good clarification, Sterling. Yes. Revenue contribution from Palladium in the cloud would be ratable, yes.
All right. Perfect. Thank you, guys.
Thank you.
Thanks.
Our next question comes from Tom Diffely from D.A. Davidson.
Yeah, good afternoon. First, love to go back to an earlier question on the memory side. I know we've seen some pricing declines in memory. I know you have really strong relationships with the customers there, but since we've seen them push out some capacity adds, delayed a few project, just curious, have you seen any delays or any slowdown in their design activity because of that?
Yeah, good question, Tom. So far, the feedback is from our partnership with some of the key customer, we don't see any delay in the design. Actually, they increase because there's a lot of new requirement they need, and I highlight the NVMe related area and the disaggregation of the storage that a hyperscale guy need. There's a lot of innovation and the new materials happening, so that you can really squeeze more bits into the memory cell. I think there's a lot of new development. We're heavily engaging with all the key players. I think we don't see the delay.
Okay. I guess that being said, would a pricing decline be good for your business in the sense that it would open up new opportunities or new use cases for memory?
Possible, hard to tell. So far, we are more related to the design activity, we see increase in the design activity. Memory is so essential on the whole big data and the whole AI machine learning. Latency is the scale out, the storage is significantly required because such a big massive of data from autonomous driving, IoT to the edge, the requirement is significantly increased. In some ways, put a lot of more pressure for the memory innovations, and that should be good for us.
Yeah. Okay. John, you talked several times about the one-time OpEx benefits in the quarter. Was there ever a discussion not to include those in the non-GAAP numbers?
No. We didn't discuss not including them.
Okay. Was it more of just a timing issue then?
That's all. Like I say, yeah, it's just a timing issue. If I look at the second half without those, we're still very pleased with how the second half of the year is playing out.
Okay. Just based on the really strong operating margin improvement this year versus the revenue growth at 9%, was that projected or predicted based on just the leverage in the model, or was that a much stronger operating margin improvement than you would have thought?
Well, approximately half of the operating margin improvement in the third quarter came from the revenue upside, and about half of it came from expense benefits primarily due to the timing of hires and those one-time credits. Those were the major impacts. Like I said, I wouldn't focus too intensely on any one quarter. Like I say, when you look at the second half of the year, we're very pleased with how that's panned out.
Okay. Thank you.
Thanks.
Thank you.
Our final question comes from Jay Vleeschhouwer from Griffin Securities.
Thank you. John, let me start with you with a couple of questions regarding pricing and separately expenses, then a follow-up for Lip-Bu. On pricing, since you assumed the role of CFO, one of the things that you've been focusing on is what you've called deal quality metrics and relatedly, improving in areas of suboptimal pricing where it exists. Could you give us an update on how you think you're doing in terms of those metrics? And then relatedly on pricing, to what extent do you think that new technology is helping you with either bookings or incremental pricing? Like for example, Voltus XP as a source of incremental business.
Jay, on the pricing front, yes, and we will continue to focus on pricing. We're always disciplined and value-driven. Typically, we get pricing improvements from add-on contracts from customers. We typically do a renewal with a customer every approximately two to three years. Throughout the duration of the contract, you'll get add-on opportunities, and that's where we often see the pricing improvement.
Just to add on to what Alan Lindstrom is saying, I think clearly, Jay, I think the customer are paying for the value that we provide. That's why we do a lot of innovations. Every year, as you can see, six to eight new product organically developed currently is driving the tool performance so that we can really focus on the value to the customer running the PPA, the runtime. In the early days, the last few years, we've been very focused on rewriting some of our tool on parallelism. The next step, we're using AI machine learning across applying into all our different tool and products so that we can drive better performance and throughput for the customer.
Lately, as we highlighted to you, we're moving quite a bit into the cloud so that we can use the unlimited server to be partitioning appropriately to scale the performance and runtime. I think clearly the customer willing to pay for the value that we provide.
Now with regard to expenses in doing our monthly spot checks of your job openings as well as for your principal competitors, across the board for the EDA big three, there has been a substantial increase over the last number of months in terms of total openings. Yourselves, Synopsys, Mentor. In your case, your current number of job openings as we've looked at today, is up by about a third from six months ago and more than double a year ago, and we see similar trends again at Synopsys and Mentor. Could you talk about where you're looking to add and given the broad demand within the industry for engineers, AEs, and the like, could you talk about your ability to, in fact, bring people on at the rate that you want given the competition for headcount?
Jay Vleeschauwer, I'll start and I'll let Lip-Bu Tan then chime in with where we're adding. Yes, you're correct that if you look at our CFO commentary, you'll see there was an uptick in Cadence headcount, and much of that from the beginning of September, which was part of the delayed hiring. That's what benefited Q3. On the expense side, that was a one-time benefit to Q3, but of course we ramped up hiring so that expense turns up in Q4. I just want to refer back to why we didn't non-GAAP out the one-time credits for professional services. That was because we need to be consistent with the use of non-GAAP definitions, and they won't change on a quarter-to-quarter basis.
Yeah, just to add on John's descriptions, I think clearly, we're hiring more into the R&D and FAE side. The FAE, clearly we want to make sure that we support our customer, and then have a deep collaboration with them. On the R&D front, we see a great opportunity for further innovations, basically more into the data science and into the machine learning, deep learning, and also some of the new tool that we have some new idea how to drive more success in terms of PPA and runtime to serve our customer better.
Lastly, on technology, since you used the word rewrite, Lip-Bu, earlier. At a Cadence customer event at DAC four months ago, there was a very interesting panel discussion where customers such as NVIDIA in particular talked about their need for substantially greater capacity in the tools. I know this is a 30-year issue in EDA, it never ends. Customers like them and others on the panel were talking about multiples of increase in block capacity within the tools, and I'm wondering where you stand in terms of being able to deliver against those kinds of substantial increases in capacity that they're now talking about.
Yeah. I think clearly the design complexities increase substantially, based on the various application I just mentioned earlier, big data and data analytics. The capacity requirement, how to address the interconnect high speed, and also some of this memory scale-out. All these are going to be critical. Besides for our tool, how to scale it, and also some of the complex design, how to use the cloud to address some of these requirements. The other part is also our emulation, hardware emulation. Clearly, we are developing the next generation. Too early to give you the guidance. So far, we are making good progress on the next products and then going to be continue to increase the capacity to meet the customer. Same thing with our FPGA prototyping. Stay tuned.
We're going to have more announcement in terms of increasing the capacity and the scale-out to provide the customer needs, like the company that you mentioned earlier, NVIDIA. Any large-scale design, like massive parallelism and AI machine learning application in some of the big infrastructure switch, network switch-related, require capacity increase. Also the developing the design is getting a lot more complex and also meanwhile pushing into the five and three nanometer. I think this is all exciting for us, and we work closely and listen very closely with the customer and collaborating with them and supporting them.
Thank you very much.
Thank you.
There are no further questions at this time.
Let me start. First of all, in closing, through continuous innovation and execution, our system design enablement strategy has positioned us to capitalize on multiple technology waves and further proliferate our solution with a broader base of customers. We are proud of the innovation and inclusive culture we are building at Cadence. I would like to take this opportunity to thank all our shareholders, customer and partners, board of directors, and hardworking employees globally for their continued support. Thank you all for joining us this afternoon.
Thank you for participating in today's Cadence third quarter 2018 earnings conference call. This concludes today's call. You may now disconnect.