Arm Holdings - Earnings Call - Q1 2012

Transcript

Moderator (participant)

Ladies and gentlemen, thank you for standing by and welcome to the Q1 Analyst Conference call. At this time, all participants are in a listen-only mode. There will be a presentation followed by a question-and-answer session, at which time, if you wish to ask a question, you'll need to press star one on your telephone. I must advise you the conference is recorded today on Tuesday, the 24th of April, 2012. I would now like to hand over to your first speaker today, Mr. Ian Thornton. Please go ahead, sir.

Ian Thornton (Head of Investor Relations)

Thank you very much indeed. Good morning, everybody. This is Ian Thornton, the Head of Investor Relations at Arm. On today's Q1 Results Conference Call, we have with me Warren East, Chief Executive Officer, and Tim Score, our Chief Financial Officer. On today's call, Tim and Warren will take us through the highlights and comments from the quarter's results, then we'll open up the call to a Q&A session. As a reminder, the presentation and release can be found on the Arm Investor Relations website at www.arm.com/ir. Before I hand over to them, I just have to read out a few words with respect to this conference call and what we're about to discuss.

The contents of this conference call are being directed only to those of you who have professional experience in matters relating to investments, and the information communicated on this call is being made available only to investment professionals. Any persons present on this call who do not have professional experience in matters related to investments should not act or rely on the contents of this call. The following conference call will contain forward-looking statements which are other than statements of historical fact. The company's actual results for future periods may differ materially from these statements as they are based on current expectations and are subject to a number of risks and uncertainties. On this note, I'll hand over to Warren.

Warren East (CEO)

Thank you, Ian. Good morning, everybody, and thank you for joining our call. I will start with some context, then go into business highlights and hand over to Tim. As usual, I hope we'll cover most of the content in Q&A. To some context, 2011 was an excellent year for Arm, and we're pleased that this morning we can report that the first quarter of 2012 has shown our momentum continuing. We have some further healthy indicators that are underpinning the long-term growth opportunity for our business. We came into the year with a record backlog, with several innovative new products in development, which will enable our customers to enter new markets and a strong competitive position in our target markets.

We're continuing to see an increase in the demand for our smarter, lower-power technology, particularly as all aspects of people's lives become enhanced with the digital products that result from that. That's driving both our licensing and our royalty revenues. In the first quarter of the year, we saw demand for our technology driven by a huge variety of end markets, from highly efficient servers and high-performance mobile computers right through to very low-power energy shipping sensors and low-cost microcontrollers. If we look at the volumes actually shipped, our royalty revenues again outperformed the overall semiconductor industry as our customers launched products into new markets and gained market share within existing target markets. This revenue growth has enabled Arm to continue to invest further in R&D, enhancing our ability to innovate and continue to develop new products.

At the same time, we've grown earnings by 23% and delivered strong cash generation. Looking forward to the rest of the year, despite Q1 industry shipments declining sequentially, most analysts expect the industry to recover in the second half. It is in that context that we expect the group dollar revenues for the full year will be in line with current market expectations. Now, I'll look at some of the business highlights and discuss the revenue drivers in the different parts of the business, starting with the processor division, where all our licenses were signed for either Cortex or Mali processors. We signed 22 licenses in the quarter. It was a broad range of end markets.

I've already mentioned servers and sensors, but we also signed licenses for product developments in security, security modules for gaming consoles, mobile phone power controllers, which of course adds an extra royalty opportunity for mobile devices, deeply embedded chips in electric motor controllers and smart meters, and of course the high-end apps processors for mobile computers and digital TVs, and for both low and high-end smartphones. If we look at where these licenses were sold, just over half the licenses were for customers based in Asia. This included Cortex-A and Mali processors for consumer electronics, Cortex-R processors for baseband modems, and Cortex-M processors for microcontrollers and smart cards. Over the years, we've increased our focus on the major economies in Asia. Last November, for instance, we announced we'd opened a design center in Hsinchu in Taiwan to deepen our collaboration with the major companies in that region.

We licensed eight Cortex-A processors, mainly to consumer electronics companies and people building chips for servers. These included another lead partner for our new Atlas product, which is based on the next generation Arm architecture, which includes 64-bit support. We also sold licenses for Cortex-A5, Cortex-A7, and Cortex-A15. We sold 10 licenses for Cortex-M processors, mainly for microcontrollers and smart cards. In the M processor family, during the quarter, we announced the Cortex-M0+, which is our smallest and lowest power processor so far. I'm particularly excited at the prospects for that processor going forward. This new technology will enable our customers to have software compatibility across their entire product range, but now including chips that were previously 8-bit microcontroller parts. The lead partners for that were NXP and Freescale. Two of the licenses sold were for Mali graphics processors into digital TVs and mobile computing.

Both of those design wins are alongside a Cortex-A family processor. Putting those together further increases our royalty percentage per chip. Talking of royalty, we'll now talk about that. Remember, our royalty revenues are reported a quarter in arrears. The numbers we're talking about, our royalty in Q1 was generated from chips sold during Q4 last year. Processor royalty revenue was up 6% year on year compared with the relevant industry revenues declining by about 2%. Continued outperforming of the industry. 1.9 billion Arm processor-based chips were reported in the quarter, and that's a 5% year-on-year increase. It was mainly driven by growth in non-mobile markets. In Q4 last year, we reported 2.2 billion chips, so there was a sequential decrease. That was primarily due to an 11% decrease in industry shipments overall.

Arm is particularly exposed to markets like disk drives and microcontrollers, and these sectors saw the largest falls. With fewer low-cost microcontrollers, the change in the mix of Arm-based chips has contributed to a sequential increase in the average royalty revenue per chip, from $0.045 in the previous quarter to $0.048 in Q1 reporting today. Part of the sequential increase was also due to an increase in the number of Cortex-A and Mali processor-based chips, especially into things like smartphones and digital TVs. Arm typically receives a higher royalty percentage per chip for chips containing Cortex-A processors and for chips with multiple processors, including Mali graphics. Over the last year, we've seen a near doubling of Cortex-A family shipments and a 30-fold increase in Mali graphics shipments. That's against the backdrop of the continuing growth of chips with two or more Arm processors.

The benefit arising from the combination of all this more Arm technology per chip is illustrated by the average royalty revenue per chip in mobile devices increasing by 5% over the past year, while the average royalty revenue per chip in home electronics has increased by 15% over the last year. Switching now to our physical IP division, total physical IP revenue was up 6% overall, $24.7 million. During the quarter, we signed licenses for a new 130 nm royalty-bearing platform, several updates, and extensions to existing platforms, spanning technology all the way from 130 nm to 28 nm. We're continuing to see demand for our processor optimization packs. These comprise physical IP optimized for use with our Cortex-A family processors. During the quarter, we signed another three POP licenses, including the first POP license for a Cortex-A7 for use in mobile and mobile computing apps processors.

The other POPs were for Cortex-A9 and Cortex-A5 processors, demonstrating demand for physical IP across the full range of Arm's apps processors. We also announced the availability of a Cortex-A15 POP on TSMC's 28 nm process and a Cortex-A9 POP on GlobalFoundries' 28 nm process. Underlying physical IP royalties in Q1 were at $11 million, which was up 9% year on year, compared to foundry revenues, which declined 3% in the relevant period. Royalty revenue from physical IP at advanced nodes, that's 45 nm and newer, continues to increase, and now that's accounting for approximately 1/3 of our total physical IP royalty revenues. Development systems had a great quarter, and we won a large one-off deal with a major semiconductor company. That is a one-off deal. Arm is continuing to transition the business to focus the tools business to focus on microcontroller tools and premium toolkits for multi-core systems.

This is against an industry context of increasing adoption of open-source tools. As we stated last quarter, due to this transition, we expect that full-year revenues for our development systems business will be broadly flat year on year. What have we been doing from an operational point of view? We've continued our investment in R&D, grown our engineering teams working on advanced processors and graphics. That was an addition of 60 people during Q1, and we expect to continue that investment in new people during Q2. The result of that investment can be seen with some of the new technology announcements I've already referred to, things like the Cortex-M0 and the new processor optimization packs. In addition, we recently announced a joint venture with Gemalto and Giesecke & Devrient to deliver improved security for applications and services running on top of our connected devices.

This joint venture will create and deliver a secure operating system, creating a so-called Trusted Execution Environment. That utilizes Arm's TrustZone technology, which can be found in all the Cortex-A family of processors. That's a quick run-through of the highlights in the business. I'll now hand over to Tim, who will provide some further detail on the numbers.

Tim Score (CFO)

Thanks, Warren. Good morning, everybody. Hopefully, most of you have had a chance to have a quick look at our Q1 earnings release out this morning. In addition to the financial information included in that release, the quarterly slide set is available on our website as usual to help with models, etc. The headlines overall: Q1 dollar revenues $209.4 million, up 13% year on year. Particularly strong growth in processor license revenues, which are up 27%, $65.2 million. That's now four sequential quarters with licensing around $60± million, which is consistent with our guidance. We expect license revenues to continue to be around $60 million in the second quarter, although it's too early in the quarter to tell if it's going to be $60+ million or $60- million. Going into Q1, order backlog was at record levels.

We exited the quarter with backlog marginally lower sequentially, and up some 15% higher than a year ago. In the first quarter, approximately 60% of PD license revenue was generated from backlog. That's at the upper end of the typical 40%-60% range. The usual analysis of backlog maturity and composition is included in the slide set on the website. That shows that 26% of total backlog is expected to be recognizable as revenue over the next two quarters. As Warren said, processor royalties were up 6% year on year to just under $93 million on the positive side of the $90± million that I referred to at our Q4 results at the end of January.

The royalty revenues were impacted to some extent in Q1 by the slowdown in Q4 of the sales of disk drive components as a result of the Thai floods, and by the general slowdown in semiconductor sales in Q4 that Warren referred to. In previous years, normal seasonality has resulted in a modest sequential decline in royalty revenues from Q1 to Q2. Industry data indicates that semiconductor sales were likely to be down about 5% in the period that's relevant for our Q2 royalties, i.e., Q1. Most analysts are forecasting that Q1 is the bottom of the current cycle, which would benefit our royalty revenues in the second half of the year. Normalized OpEx in Q1 was GBP 66.1 million compared to the guidance given in January of GBP 64 million-GBP 66 million.

As noted in the earnings release, the increase in OpEx over the last 12 months is due to increased investment in our research and development teams. Normalized OpEx for Q2 is expected to be in the range of GBP 67 million-GBP 69 million, assuming current exchange rates. Sequential increase is expected as we continue to invest in the development and deployment of new technology. Normalized operating margins were over 44% in Q1 this year compared to 43% a year ago. Cash generation continued to be strong, demonstrating that we're managing to increase profitability and generate good levels of cash whilst increasing investment in future product development and organizational capability. Now, onto the outlook. Both the order backlog and the licensing pipeline are robust, pointing to continued strength in processor licensing revenue. Arm Q2 royalty revenues are generated, as I say, from Q1 chip shipments.

As mentioned earlier, relevant data for the first quarter indicates that industry-wide revenue is down about 5%. With most analysts expecting recovery in the second half, in that context, Arm expects group dollar revenues for the full year 2012 to be in line with current market expectations, which are $879 million. That is just under $20 million higher than the full-year consensus revenues that were in place at the end of January when we did our full-year results. Now over to Q&A.

Moderator (participant)

Thank you, (Inaudible). We will now begin the question-and-answer session. If you wish to ask a question, please press star one on your telephone keypad and wait for a name to be announced. If you wish to cancel a request, it's the hash key. That's star one to ask a question. Your first question comes to the line of Vivek Arya from Merrill Lynch. Please ask your question.

Vivek Arya (Managing Director and Senior Analyst)

Thanks for taking my question. I'm curious what you see in the industry regarding the supply issues at some of these advanced nodes, 28 nm specifically. Do you think that'll have an impact on your second half results?

Warren East (CEO)

Hi, it's Warren. We're not anticipating an impact. I think you're probably referring to the comments made by Qualcomm. From what we can see, Qualcomm are experiencing a very strong demand for some particular designs. They highlighted that it would be a supply-constrained situation rather than a demand-constrained situation as far as those particular products are concerned. Don't forget, Qualcomm is just one of Arm's customers. Many of our semiconductor partners are seeing demand for Arm-based technology. We don't think that that will be a material impact on our business.

Vivek Arya (Managing Director and Senior Analyst)

Got it. Just on a separate subject, when is the earliest, Warren, that we should start thinking about impact from Windows 8? I know it's probably very small this year, but how should we think about the impact on your revenues? Is it more a second half 2013 impact, or how should we think about it?

Warren East (CEO)

I think a few things to take into account. First of all, the timescales are very heavily controlled by Microsoft and not by Arm or Arm partners. Secondly, you're right not to get any serious expectations for the current year because certainly all indications are that product launches are right towards the back end of this year. Thirdly, these types of products, you know, we're talking about things like mobile computers and tablets and the like, there is a little bit of a bias towards consumer, and that does tend to be a little bit seasonal. That would point towards later in the following year as being time for more demand. The story as far as we're concerned is that this is a long-term thing. I know it's very exciting as far as you know when is the impact actually going to hit.

We see this as a continuous opportunity over several years. Exactly when it starts, it's going to be only really noticeable after it's happened.

Vivek Arya (Managing Director and Senior Analyst)

Thank you very much.

Warren East (CEO)

Thank you.

Moderator (participant)

Thank you. Your next question comes from the line of Sandeep Deshpande from JPMorgan. Please ask your question.

Sandeep Deshpande (Subject Matter Expert)

Yeah, hi. Thanks. Firstly, Tim, could you talk about this increase in OpEx? Why is this OpEx increasing more than we were expecting? You talked about R&D, but exactly what is changing in your R&D spending trends? Secondly, if we could talk about the increase in your Cortex-M shipments. Cortex-M seems to be increasing quite substantially as a percentage of your shipments. At the same time, your royalty per device also seemed to have increased, which seems to be a factor from Cortex-A. Maybe we can talk, if you could just talk about these dynamics within your royalty mix.

Tim Score (CFO)

Yeah. Hi, Sandeep. On the OpEx runtime, as I said in the comments, guidance for Q1 was GBP 64 million-GBP 66 million. It came out at GBP 66 million. Consensus was GBP 65 million. GBP 1 million, I don't think it's particularly significant on a quarterly basis. What we are doing, as we've said, is investing in our R&D capability and our technology roadmap. That's in processors, in graphics, in physical IP, and also in our organizational capability as we grow. I don't think there's any particular surprise there in OpEx. We do see that we will continue to invest, subject to the market staying or improving in the second half.

Warren East (CEO)

Okay. I'll talk to the Cortex-M question, Sandeep. Cortex-M is obviously our processors aimed at microcontroller products. We are seeing very strong growth there. Year on year, we saw microcontrollers grow 35% this quarter with industry shipments actually down 10%. That is Cortex-M-based products taking market share. If you look at it on just a quarter-on-quarter basis, microcontroller units, Arm-based microcontroller units were down 20% compared with mobile devices down 10%. The Cortex-A products are typically commanding a higher royalty rate, and they're going into higher-priced chips. We saw the increase in average royalty per chip driven by the relative higher shipments of Cortex-A compared with Cortex-M. That's what drove the average royalty up. Cortex-M-based products, we expect, as you look sort of on a longer-term basis, as I said, year on year, it is significant industry outperformance. We expect that to continue.

At the end of 2011, our market share was approximately 10% in microcontrollers. We do see that continuing to grow over the years. Cortex-M is a serious driver. As I mentioned in the introduction, Cortex-M0+ will help accelerate that by really opening up the opportunity to replace very small, very low-power 8-bit microcontrollers.

Sandeep Deshpande (Subject Matter Expert)

Just to follow on, Warren, you talked about 10% in the microcontroller market in 2011. Do you have any metric to help us understand whether there are about 50 microcontroller companies that you've targeted to license your IP to and whether you've succeeded in licensing it out to 40 or anything like that?

Warren East (CEO)

That's a good point. There are hundreds of these, of course. A few years ago, we produced a slide in our earnings pack where we were looking at semiconductor companies producing microcontrollers, those of whom who had licensed Cortex-M and those of whom who were using Arm in their microcontrollers. We haven't updated that slide in this set. Perhaps we should, looking forward, maybe include some information on that over the coming weeks. You know, the opportunity is very large because some of these companies are shipping products based on an 8-bit processor, and it's just into very niche applications. There are literally hundreds of them. We've sold well over 120 Cortex-M licenses so far, and we expect that to continue to grow.

Tim Score (CFO)

Yeah, and I think it's worth saying that, I mean, this is Tim. You know, the royalties that we earn from microcontrollers today are earned from a relatively modest number of semiconductor partners, you know, around about 10-20. As Warren said, in the last three or four years, we've signed just about 140 Cortex-M licenses to over 100 companies. You can see that the vast majority of those have not yet come to market with Arm-based products but are all planning to do so.

Sandeep Deshpande (Subject Matter Expert)

Thank you.

Tim Score (CFO)

Thanks.

Moderator (participant)

Thank you. Your next question comes from the line of Gunnar Plagge from Citi. Please ask your question.

Gunnar Plagge (Research Analyst)

Yes, hello. Mobile unit growth in the first quarter seems to be down by about 4%. I was just wondering whether you could give us a bit of an idea of what you expect next quarter and where we should probably end up for the full year.

Warren East (CEO)

Obviously, as we look at our royalties that we're going to report next quarter, there's going to be a very strong correlation with industry shipments in Q1. As we have a very, very high share in the mobile space, then Arm's shipments there will reflect the industry. To that extent, when we look at the prospects for the mobile phone space overall, and in particular smartphones, analysts are across the board, you know, generally, we're looking at about a 35% increase in smartphones in volume year on year. That should be reflected in Arm's numbers. We are seeing trends for sort of integration of functions, so combinations of sort of Wi-Fi and Bluetooth and that sort of thing, which will, when you look at it in volume terms, mean the numbers deviate slightly from that that I just outlined.

Generally, our mobile shipments will be driven by that strong 35% growth in smartphones that people still expect this year.

Gunnar Plagge (Research Analyst)

Okay, thanks. As a follow-up, you launched the A15 hard macro for TSMC. I remember about three years ago, we talked about the Osprey and with regard to that hard macro, you said, "Once in a while, we do that, but I don't see a substantial industry trend." Now, on the back of substantial competitive pressure vis-à-vis Intel, do you think there's an overall trend for maybe a diversification of system-on-chip design strategies away from RTL and more to maybe potentially hard macros?

Warren East (CEO)

It's certainly a possibility. You know, we keep in very close touch with the trends and the desires of our semiconductor partners and those people who use foundries like TSMC. I think occasionally you will see the hard macro like we did Osprey. A more exciting trend for us is the processor optimization packs where we can optimize our physical IP and give people the flexibility of RTL-based design combined with the performance power characteristics and the sort of optimization characteristics that you get from designing to the transistor level. That's what our processor optimization packs are all about. In Q1 this year, we saw continued demand for those. The reason that we launched the ones that I mentioned, the new ones, is based on dialogue with our partners and their foundry partners. That's based on future demand that we can see.

Gunnar Plagge (Research Analyst)

Okay, thanks.

Warren East (CEO)

Thank you.

Moderator (participant)

Thank you. Your next question comes from the line of Francois Meunier from Morgan Stanley. Please ask your question.

Francois Meunier (Head of Technology Equity Research)

Yes, thanks for taking my call. Just a quick question on the tax rate, maybe longer term. We've seen in the U.K. budget that there was a Patent Box and potentially the tax rate could decrease significantly for Arm in the coming years. If you have a bit on that. The two other questions I have, first is on servers. You signed a 64-bit license with a lead partner. Could you confirm this is a relatively large customer? Do you feel more upbeat about server following signing this license? My third question, if I may, is about the royalty rate you got this quarter. I know you don't disclose exactly this number, but was it up year on year to compare to Q1 2011? Thank you.

Tim Score (CFO)

Yeah, I'll deal with the tax. As you say, Francois, the Patent Box is moving towards its introduction. It's going to be introduced in April next year, 2013. There will be, if you like, a transitional implementation where a significant proportion of the benefit will be introduced in year one, and the remainder will be introduced gradually over the next few years. Overall, obviously, the devil is in the detail of what constitutes a qualifying patent, etc. Certainly, by the time we're through the implementation period, you can expect a significant reduction in Arm's effective tax rate. This year, this quarter, you've seen us looking at a 25% normalized rate for this year, and the expectation would be post the introduction of Patent Box that the rate will be potentially sub-20%.

Francois Meunier (Head of Technology Equity Research)

Sub-20%?

Tim Score (CFO)

Yeah.

Francois Meunier (Head of Technology Equity Research)

Okay, very good.

Tim Score (CFO)

The other thing to bear in mind, which is also not yet definitive but is likely to occur, is the reclassification of the R&D tax credit from the tax line to above the line as an offset against operating expenses. That is neutral to earnings. When considering impact on tax rate, bear in mind that if it is introduced, and if it is introduced, that's expected to come in around the same time as Patent Box, that would increase operating margins but also increase tax rate with neutral on earnings. That's just also worth bearing in mind should that be introduced. More detail to follow as and when.

Warren East (CEO)

Your other questions, Francois, the question about the server opportunities. You have two licenses there, and we put in the earnings release. One is driving a new lead partner for Atlas, and one is a Cortex-A15. Obviously, these are high-end processors. They're high-value products from Arm and therefore more likely to be sold to larger semiconductor companies with more resource to take on those sorts of very large-scale development. We're certainly very enthusiastic about them. As for the customers involved, they will be sort of revealing themselves as and when they want to make announcements about their products. We can't really comment any further on those customers now. We're quite pleased with those design wins. I think it would be very helpful for Arm's activity in servers. Your question about royalties, typically, we don't really get involved much in talking about trends year on year on this particular parameter.

As you know, there's a combination of upward pressure on the royalty rates caused by multi-core and higher-value Arm processors. There's a downward pressure caused by increasing success in the microcontrollers space, which tend to be lower-priced chips. The result is a combination of those two factors. As I said earlier, for this quarter, it happens to have resulted in an uptick in our royalties. We absolutely expect the upward pressure from more higher-value chips to continue. We welcome market success in the microcontrollers space as well. That's why we don't really forecast what the number is going to be for the whole year.

Francois Meunier (Head of Technology Equity Research)

Thank you.

Moderator (participant)

Thank you. Your next question comes from the line of Gareth Jenkins from UBS. Please ask your question.

Gareth Jenkins (Head of Research)

Thanks. Yeah, Gareth Jenkins. Just a couple of follow-ons. One from Francois. I wondered if you could give a sense of whether die size increases have been a driver in the last couple of quarters. Obviously, we've seen some fairly sizable increases on the apps processing side. Secondly, I just wondered now we're a bit of a way into the year, whether you feel comfortable giving us more of a steer in terms of Mali units and the run rate at the end of this year that you're expecting. I think last year you gave a very helpful steer on that. Finally, Warren, just on big.LITTLE, what the development is in terms of licensing and progress with the big.LITTLE development. Thank you.

Warren East (CEO)

Okay. First of all, let's do die size. Obviously, for processor royalties, it's really driven by chip prices. Whether there happens to be a big chip or a little chip, it's the price that the chip is sold for that really matters. Clearly, semiconductor companies are going to try to sell the chips that cost them a lot more at higher prices. To that extent, big applications processors with multiple processors, lots of memory, graphics processors, and other hardware accelerators pushes the die size up, pushes the chip price up. With our percentage-based royalties, it's one of the contributors for driving up the price per chip. In terms of die size specifically and where it might make a difference to Arm, it's probably in our physical IP royalties. You'll note that we said 1/3 of our physical IP royalties are now derived from 45 nm and newer.

As that trend continues, we will see more of our physical IP royalties driven by apps processors. To that extent, we should see upward pressure from the increasing die sizes because the royalty in our physical IP business is on a per-wafer basis. Fewer chips per wafer means more wafers means more royalty. On Mali, we expect to see ourselves growing through this year. We said that last year was going to be a year of tens of millions of units in 2011. Indeed, it was. We're hoping that this year we'll be breaking the 100 million barrier for the full year as a whole. That's where we expect on Mali. Your third question, Gareth, was on big.LITTLE. We have 11 partners committed to going down the big.LITTLE route at the moment.

We are in the phase where the first chips are sort of being manufactured at the moment. We're looking forward to silicon later in the year. That will be the next milestone for the big.LITTLE story. It's going pretty well. As we've said before, we think big.LITTLE is going to be a serious step forward in terms of power efficiency for the Arm story as a whole.

Gareth Jenkins (Head of Research)

Thank you.

Moderator (participant)

The next question comes from the line of Andrew Gardiner from Barclays. Please ask your question.

Andrew Gardiner (Research Analyst)

Thanks very much. Firstly, I was interested if you could shed a little bit more light on Windows on Arm. I know the last time the questions came up after fourth quarter results, Microsoft hadn't given a whole lot of detail around the project. Now they're steadily giving more detail about the pending launch, how they're viewing Windows on Arm or rather what they're calling Windows RT now. How are you seeing your semiconductor partners react to this or plan for the pending launch? In particular, on the software porting side of things, do you have anything you can share on that? Also, just a quick two follow-ups on the server side. Can you tell us, obviously without naming the company, whether they're existing or new customers? Finally, on PIPD, when do you expect to see the POP royalty revenue to start in more significant volume?

Warren East (CEO)

Okay. Let's look at Windows on Arm to start with. This is both an engineering program and a product launch. Microsoft are managing the product launch, and they are obviously a key player in the engineering program. Our semiconductor partners who are involved are also key players in the engineering program. As far as we can see, that program is on track and aligned with Microsoft's marketing launch. There isn't really any new news on Windows on Arm. All the blogs which came out during Q1, the blogs which have emerged over recent weeks, are milestones on the journey towards both the engineering program getting to fruition and the marketing launch. There is no other sensible thing that I can add to it. You need to talk to Microsoft if you want to talk any further detail about that.

Software porting for the chips involved is part of the engineering program, and that is underway. The OEM customers who are being supported are being supported. As far as servers are concerned and the server announcement, there are existing partners. As I said, there will be some individual product launches as and when those partners want to talk about them. We actually have already seen our first royalties for the POPs. We saw that in the fourth quarter of 2011. We would expect, given the trajectory of licensing of POPs over the last 18 months or so, to see that gradually increase. It's a component of our physical IP royalties. Bear in mind that the time to market here is similar to the time to market from when we license a microprocessor core. Typically, to get into appreciable volume, we've said that's a four-year exercise.

Although we've seen the leading edge of that, it's really a very minor contributor at the moment. That should grow significantly over the next several years.

Andrew Gardiner (Research Analyst)

Thanks very much.

Moderator (participant)

Thank you. Your next question comes from the line of Simon Schafer from Goldman Sachs. Please ask your question.

Simon Schafer (Head of European Technology Research)

Yes, thanks so much. Actually, Tim, I wanted to ask a question on OpEx growth. I think, you know, historically, you've always sort of hinted towards the expectation perhaps that sales growth should be outpacing OpEx growth by a factor of two to one or something to that effect. This year, I think it sounds like perhaps your OpEx growth is significantly higher than that ratio. I guess, you know, post this year, which is still a big investment year in a couple of the peripheral technologies and so on and so forth, do you expect that OpEx growth to perhaps have a lower run rate in comparison to your revenue growth opportunity?

Tim Score (CFO)

Yeah, hi Simon. I think there's no change to the overall messaging that we give here around the shape of the Arm business model and how we expect it to develop financially over time. The operating leverage is there. We expect over time royalties to grow faster than licensing. We expect costs to grow more in line with licensing than with royalties. Therefore, you're going to get margin expansion. Looking at it quarter to quarter and even year to year doesn't really give you the full picture. If you look at last year, obviously, revenues grew well ahead of costs. We'll see exactly how that develops this year. The basic position is there's no fundamental change to how we are seeing the development of operating expenses. There's no change in what we feel we need to do to achieve the growth.

Therefore, the margin expansion that we've seen in the last two to three years, from 30% into the mid-40%, we continue to see the margin expanding. As I said before, if you look forward four or five years' time, is the margin 50%? Is it 55%? Is it 60%? That will depend on what we're investing in at the time for future technology development and license revenue opportunity. Of course, it will also depend on precisely how our market penetration grows in target markets. The basic message is no change to the relationship between OpEx development and revenue growth in this business.

Simon Schafer (Head of European Technology Research)

Got it. Thanks, Tim. Thank you.

Moderator (participant)

Your next question comes from the line of Nick James from Numis. Please ask your question.

Nick James (Equity Research Analyst)

Yes, good morning. I just had a question follow-up on PIPD in terms of licensing revenue. Obviously, there's a lot of kind of qualitative signs of momentum with the POPs and new nodes. The licensing revenue seems to be kind of stuck around this kind of $11 million-$12 million a quarter. Are we expecting it to kind of grow from this level at some point? If so, when?

Warren East (CEO)

We do expect it to grow. It's going to grow modestly. Part of that growth is because of an increasing shift towards newer technologies, leading-edge processors, and that type of thing. You'll notice I talked about a whole range of licensing done during the quarter, ranging from 130 nm to 28 nm. A couple of quarters ago on a call like this, I would have been talking about a range from 250 nm to 28 nm. The bias is shifting more towards the newer technologies. Those newer technologies take more effort. They're higher-value licenses, and we'd expect to see that revenue come through. Because there is a lot of new technology in there, a lot of the orders that we take for physical IP licensing actually go into the backlog. The license revenue that's actually recognized takes longer for these newer technologies simply because the development takes longer.

Nick James (Equity Research Analyst)

Great. Thanks very much.

Moderator (participant)

Thank you. Your next question comes from the line of Janardan Menon from Liberum Capital. Please ask your question.

Janardan Menon (Technology Analyst)

Hi, thanks. I was just wondering, Tim, maybe if you could give us a feel for how you expect licensing, especially BD licensing, to move over the next few quarters. You've guided down licensing again to around the $60 million level. One, would that be feeding off the backlog again? Would you expect to replenish the backlog? And two, by around what quarter, based on your visibility right now, would you expect licensing revenues to start inflecting back up again on a growth trajectory?

Tim Score (CFO)

I think it's worth putting that question in context. I mean, our long-term processor license revenue growth guidance is mid to high single digits. I think we've seen in the last two full years, 2010 and 2011, growth well in excess of that. 2011, bouncing out of 2010, bouncing out of the downturn. 2011, strong. Both 30%+ years. In Q1, as you've seen, growth of 27%. We're actually not guiding down license revenue. We're actually maintaining consistent guidance of $60± million. In some quarters, it'll be plus as it has been in the last two quarters. In some, it potentially will be minus. In terms of an order backlog that is, although marginally down this quarter, still at record levels and more than 2x where it was two years ago, that's a very strong underpin for future license growth.

As I said in the opening remarks, the opportunity pipeline is significant. We do see continued strength in licensing. If you look at over multiple periods, even off this much higher base that has been created over the last couple of years with significantly above-trend growth, we are pointing to license revenue growth continuing to grow in the sort of mid to high single digits as we go forward from this higher base. That guidance remains unchanged. The quarterly fluctuations are obviously, there's still potentially some lumpiness there, which is why we say $60± million rather than a more definitive number.

Janardan Menon (Technology Analyst)

Okay, got it. I'm just a bit curious on a small comment in the press release, saying that the average royalty revenue for mobile devices increased by 5%, while for home electronics, it increased by 15%. I was just wondering, what are the dynamics in home electronics which is driving it up? Is it just a shift towards Cortex-A class processors which is behind that strength?

Warren East (CEO)

The shift to Cortex-A processors applies to our growth in the home electronics as well as in mobiles. A key difference is the penetration there in Mali graphics, where in things like digital TVs, we have a very strong share in Mali graphics. That's certainly helpful. The other thing to note, of course, is that mobile is much bigger than just the application processor piece. That comprises things like baseband modems, Bluetooth, Wi-Fi, GPSes, and all those sorts of things, other things which go into mobile in a much bigger way than we're talking about with home electronics.

Janardan Menon (Technology Analyst)

Okay, got it. Thank you very much.

Moderator (participant)

Thank you. Your next question comes from the line of Paul Morland from Peel Hunt. Please ask your question.

Paul Morland (Technology Analyst)

Oh, hi. A couple of things. First of all, I just wanted to clarify that comment Tim made on tax. I thought the consensus for next year on the tax rate was about 25%, 26%. Are we now saying we should be using 20% based on the patent relief?

Tim Score (CFO)

When you say next year, are you referring to 2013?

Paul Morland (Technology Analyst)

Yes.

Tim Score (CFO)

What I'm saying, what I said was that my expectation for this year is around about 25%. What I'm saying is that when the Patent Box is introduced in April 2013, there are going to be some transitional implementation rules. As I said, the devil will be in the detail about what precisely is a qualifying patent for this tax. Our expectation is that when it is fully implemented, the rate will be sub-20%. Now, between now and when it's fully implemented, the rate is going to be on a decreasing trend, partly because of the transitional implementation rules and partly, of course, because of the already legislated reductions in the U.K. corporate tax rates. The exact line of the descending rate will be a detail, but it's going from 25% to sub-20% over the next five years.

Paul Morland (Technology Analyst)

For [DCF] models, etc., we should be using sort of 20% or below?

Tim Score (CFO)

I think a gradual step down between now and five years' time with sub-20% in five years' time. Exactly how you step down from 25% to 20% will be tied up with other assumptions and the implementation detail.

Paul Morland (Technology Analyst)

Okay, thanks. The other question was just on the cash generation in a quarter. It was a bit lower than I expected. Normalized cash generation GBP 58 million, down on Q1. I had expected that to be a bit stronger based on the high debtors at the end of last year. I just wondered what other factors were in there because I know debtors came down, and it was reflected in much better DSOs at the end of this quarter compared to the end of December. What else was in there that stopped the cash generation being better than that GBP 58 million?

Tim Score (CFO)

I think the problem is, Paul, when you're looking at quarterly cash generation, it's a pretty short period. I mean, GBP 58 million is one of the higher cash generation quarters we've had in history. It happens not to necessarily be the absolute highest, but there's nothing specific that I would want to point to. That's a higher than average quarter and consistent with where we'd expect it to go.

Warren East (CEO)

Last quarter as well.

Tim Score (CFO)

And higher than last quarter, yeah, for Q1.

Paul Morland (Technology Analyst)

I know, the last quarter was particularly low because you signed those deals at the end of December, didn't you, where you didn't get the cash? I thought with the cash coming in for those, that would have helped this quarter. If you say that there's nothing unexpected in there, then that's fine.

Tim Score (CFO)

No, I mean, I think it's the detailed flows of receivables and payables and the timing of cash payments. I think to look at it on a quarterly basis is probably slightly overanalyzing it. If you look at the trend line, the GBP 58 million is the highest quarter for a year, and you know, in the range of what I would expect as a sort of an average quarterly cash flow as the cash generation grows over time.

Paul Morland (Technology Analyst)

Okay, thanks.

Moderator (participant)

Thank you. Your next question comes from the line of Johannes Schaller from Deutsche Bank. Please ask your question.

Johannes Schaller (Technology Equity Analyst)

Yeah, thank you for taking my question. I just really wanted to get a bit of detail on that JV announcement with Gemalto and G&D. If you could just generally share a bit of your expectations for this JV over the next years and also what it does potentially do to your ASPs per chip or kind of what the revenue opportunity is from those TrustZone and TEE add-ons to the chips. Also, maybe how the revenue recognition exactly works here. Is that like an option that ships automatically and can then be activated? I think that's how I understand it at the moment. Thank you.

Warren East (CEO)

You probably need to have a more detailed discussion with us about that than we have time for on the call this morning. The JV is a three-way JV because when we looked at the software environment around our TrustZone, then both Gemalto and Giesecke & Devrient are sort of leaders in the secure operating system and the Trusted Execution Environment that sits around that. Both of them are working towards developing secure services that sit on top of that. The situation was a little bit diverse because we have sort of people developing slightly different approaches to the same thing. What we needed to do was come up with something which unified these approaches to enable both to thrive and to develop the business further.

In fact, when we did calls with journalists on announcements of the JV, we were drawing an analogy to a few years ago when people were developing different approaches to high-definition video storage. A group of people got together and said, "The answer is Blu-ray." At that point, that enabled Blu-ray to develop as a standard because it was a critical mass around one thing. That is what we're aiming to create with a joint venture. From a revenue opportunity, the main revenue will be for services provided on top. The joint venture itself will be securing revenue as and when services are activated on top of the Trusted Execution Environment. There are a number of different approaches to the business model which have yet to be developed by the JV.

That is another reason why it's a JV because, again, we needed to create a single unified revenue model in this space. As far as expertise is concerned, all three companies have slightly different pieces of the jigsaw puzzle. We felt the best way of taking those contributions to market was to get together and form a single entity which would develop this section of the market.

Johannes Schaller (Technology Equity Analyst)

That's very helpful. In terms of revenue split, whenever there are any revenues occurring, would that be split 40-30-30 for all the revenues occurring in that JV?

Warren East (CEO)

Everything to do with the JV goes into the JV. The JV owners have to account for the contributions from the JV as per their individual accounting standards.

Johannes Schaller (Technology Equity Analyst)

Okay, that's very helpful. Thank you.

Warren East (CEO)

Thank you.

Moderator (participant)

Your next question comes from the line of Lee Simpson from Jefferies. Please ask your question.

Lee Simpson (Managing Director and Equity Research Analyst)

Hi, good morning, gentlemen. I just want to ask another question on Mali if I could. I wonder if you could categorize for us really what the Mali graphics growth profile would be if we exclude Samsung LSI and perhaps Mstar and digital TVs. Adjunct to that really, how do we actually see Arm optimize their GPU to sit next to their CPU cores? I mean, how does that manifest? If we look at this possibility, I can't see this changing unless you overhaul the fabric and maybe make some changes in the memory interface. Is that the idea at hand?

Warren East (CEO)

Certainly, there are commonalities or there are technical synergies to exploit in sort of having the processor and the graphics processor working together, which indeed we intend to exploit in the product roadmap. We are exploiting in the product roadmap going forward. The companies to which you're referred to, of course, are existing licensees. They've licensed stuff which is essentially designed so they're historic from a point of view of realizing technical synergy. Actually, they don't realize a lot of technical synergy. The technical synergy piece is something for those customers and others going forward. Those customers that you mentioned, they're, of course, significant players in the target markets that shift volume. That's what's driving volume today and the volume growth that I referred to for 2012. We've sold 60-odd licenses. There are other customers as well as the ones that you mentioned.

Certainly, those are in themselves customers driving significant volume and representative of others driving volume in those types of target applications. There are some others that are slightly sort of smaller applications.

Lee Simpson (Managing Director and Equity Research Analyst)

Okay. Maybe just a follow-up question if I could. I mean, if we look at the back half of the year, one of the main events happening as far as we see is the release of Jelly Bean. It looks like the first mainline release on Android that sees concurrent Arm and Intel readiness. I wonder if you could characterize for us what sort of impacts from this could there be for Cortex-A volumes in the back half and possibly into 2013, just really given that potential Intel presence that we could see?

Warren East (CEO)

Of course, the next release of Android is just one of the components that is driving the 35% or so increase in smartphones expected this year. To that extent, it's hard to single out one release of Android. It's been well documented that Intel has intent to capture a share of the smartphone space. They have a couple of customers announced so far. Obviously, having day one support on a new Android release is very helpful for that. I don't think it's on its own going to be something that we can quantify where we can quantify an impact on Intel's market share, which in this time period is, of course, likely to be pretty small anyway.

Lee Simpson (Managing Director and Equity Research Analyst)

Great, thanks.

Ian Thornton (Head of Investor Relations)

Yeah, I think we'll have time for one more question, please.

Moderator (participant)

Thank you. We do have one more question in the queue for you, sir. That's the last question from Martin O'Sullivan from Cenkos. Please ask your question.

Martin O'Sullivan (Analyst)

Yes, good morning. Thanks. Just to follow on Mali, the 100 million target, I gather digital TVs are a key portal for Mali at the moment. Can you indicate approximately the sort of mix shift you would expect during the remainder of the year as you move towards that 100 million target?

Warren East (CEO)

I mean, not.

Tim Score (CFO)

The vast majority of that 100 million would still be expecting to come from smartphones because we're only expecting a relatively of the 200 million digital TVs sold next year, probably only about a quarter of those to have any sort of 3D graphics acceleration at all. Although we do expect to have the large majority share of those of the 3D graphics in things like digital TV, that still then means that the majority of the Mali shipments this year would be in smartphones.

Martin O'Sullivan (Analyst)

Okay, thanks very much.

Warren East (CEO)

Thank you.

Moderator (participant)

Thank you. That was your final question, sir. Back to you.

Warren East (CEO)

Good. Thank you very much, everybody, for joining the call. We'll be back at the end of July with our half-year results. We look forward to talking to you then and individually in the meantime.

Moderator (participant)

Thank you, ladies and gentlemen. That does conclude your conference for today. Thank you all for participating.