Imagination Technology Group PLC (IGNMF) CEO Andrew Heath on Q2 2017 Results - Earnings Call Transcript
Imagination Technology Group PLC (OTCPK:IGNMF) Q2 2017 Earnings Conference Call December 6, 2016 4:30 AM ET
Executives
Andrew Heath – Chief Executive Officer
Guy Millward – Chief Financial Officer
Analysts
Nick James – Numis Securities
Sandeep Deshpande – JPMorgan Cazenove
Janardan Menon – Liberum Capital
Vijay Anand – Mirabaud Securities
Doug Smith – Agency Partners
Achal Sultania – Credit Suisse
Adam Bowler – Deutsche Bank
Andrew Heath
Hi. Good morning. Welcome to Imagination Technology's Interim Results Presentation. I'm Andrew Heath, CEO and I'm joined today by Guy Millward, our CFO. I guess before we start to talk about our interim results, I just like to mention, today we have announced that Peter Hill will be joining our Board to replace Bert Nordberg as Chairman and he would join us in the New Year. I would like to say that Bert has provided fabulous support to both Guy and I through all of the restructuring that was implemented during the year.
Bert, he personally felt that with his other commitments that he really didn't have the time to continue as Chairman of Imagination, and therefore he informed the Board earlier in the year that he was intending to step down. So, over the last few months have been running, you know a usual rigorous in order process to select a new Chairman. And we've selected Peter Hill as the best candidate.
I know Peter well and I look forward to working with him again. So, with that let's turn to our interim results and Guy and I will take you through the usual format. I will give you some sort of opening remarks, Guy will take you through the financial. I'll come and then talk more about the business and where we're taking the business.
So, firstly, I'd like to say we've made really good progress over the past six months, especially during a period of a busy restructuring. We are on track to put Imagination back on a sound financial footing and delivering future profitable growth. The underlying trading performance of the continuing business in the first half was in line with our expectations and we additionally benefited from the strength of the U.S. dollar.
Restructuring that we announced earlier this year back in February and March has delivered to plan and is nearing completion. And as such, the £27.5 million worth of cost savings that we announced back in February and March have been delivered and delivered in full, and we're getting the benefit of that in our first half results and will continue to do so in the second half of this year.
Importantly, the continuing business is profitable and cash generative and we've provided those numbers in our press release. And also importantly the profitability has improved over the same period compared to last year.
And you'll see in Guy's presentation that we've provided, further transparency on all of our business unit, kind of our MIPS and Ensigma, and you will see that MIPS and Ensigma have made positive progress in particularly over the past six months in terms of their profitability. And for Ensigma, particularly, they had a very good first half in terms of license sales and for the first time we've covered their R&D cost from their revenues. So, that's a good milestone achieved from Ensigma.
As we look to the full year, we remain on course to deliver an underlying trading performance for our continuing operations in line with our expectations, and we do anticipate that we will continue to see an additional benefit from the strength of the U.S. dollar.
I talked about the operational review when I was here back in July, I said at that time that it was nearing completion. We have completed the operational review. We now have a clear refocus strategy for Imagination and also the confidence that we'll deliver profitable growth from our core businesses PowerVR, MIPS and Ensigma going forward.
Now, we've finished the restructuring. We can absolutely focus on exploiting our leading technology across a range of increasing opportunities exciting opportunities and concentrate on executing our refreshed strategy. As I said, when I joined back in February, we've been very deliberate regarding capital allocation and making sure we deliver shareholder value as we go forward.
But maybe, just before I pass over to Guy, I'd just like to sort of highlight a couple of points about the restructuring. Yes, we've sold a number of businesses. Yes, we've reduced our headcount, we've already reduced our headcount by 500 by the time we finish the restructuring. But I just want to be clear that the core businesses are unaffected by all that restructuring. We have been recruiting in PowerVR and MIPS and Ensigma all through the year and that will carry on. And really now we have completed the disposals, we've virtually finished the restructuring, you can now freeze up more management time to go and focus on the core businesses and executing the strategy.
And as I said in particular, you know if we go back to February, we said that we would recruit 50 people, more people in the PowerVR. We're well through that, and actually we are aiming to recruit even more people into PowerVR in particular over the coming months to basically go and access the opportunities around there in front of us, which I will talk about later.
So, with that, I'll let Guy take you through the financials.
Guy Millward
Okay. Good morning everybody. Thank you. Many of you would have read some of this already, but I'm following on from Andrew's comments, our main message this morning is that, we think we're on track, so there's a mix, there always is in the numbers, but generally underlying when you strip out exchange we are where we expected to be at the half year and I'll go into some of the details on royalty to licensing et cetera as we go.
Some of the mix is up here, so licenses is better than it has been in recent months and in the past, but it's still a little bit slow. It could be in a slightly better position and we see some of the pipeline moving into the second half, but the deal is not going away, so there is still the opportunity to deliver that in the second half.
Then royalties in line with forecast and I'll go into some of mix that's going on in there. As we go through, that helps the profit, obviously the FX has been a big impact on that, but also the costs are reduced, and that cost reduction is baked into these numbers and is baked into the second half. So, we – that sort of done now and you'll see that come through at the kind of total level that we expected for the year.
Again, there is a bit of exchange in the cost numbers, so the cost numbers underlying are actually slightly better than you see here, it's probably about £4 million in the cost for exchange due to the fact that sort of 40% of our cost base is U.S. dollar based in either salaries or licensing arrangement.
Disposal is nearly complete. They've had – the most significant impact they've had in this half is on cash, so you will see a cash out flow from operations in the half and all of that is caused by the continued – the discontinued operations, sorry. In Pure, it didn't leave until the end of September which is most the half and IMG Works which is still there and the detail in the numbers in the announcement, but basically there is a £10.5 million outflow from those discontinued operations in the first half and that offsets the inflow from the continuing operations.
So, we're keeping the tight control of it and you'll see the net debt has gone up most of that increase is due to the retranslation of the dollar debt that's in there and we'll come to that on the cash flow statement.
Moving onto the Group P&L, getting into some of the detailed numbers here. The licensing number is up 6% to 7% year-on-year at £11.8 million. That's actually a bigger increase if you allow for the fact that in the first half last year there is about £2.4 million of revenue that we subsequently reversed in the second half of the year. I think when we were here five or six months ago, we talk about some contract we renegotiated, that revenue or some of that revenue was booked in that £11 million last year. We took that out in the second half to come to the full year numbers. And so the growth is higher. It's about 37% if you allow for that. Exchange is probably about 14% of that 30, so we're talking about 20% license growth. So, it is improving year-on-year. As I say, it could have been more spectacular, but it's a solid movement and we've got plenty of pipeline for the second half.
The royalties, I've got the shipment volumes on the slide here, and say a couple of words about that. Firstly, MIPS is up 11%, so there is that solid growth across the customer base. There isn't a customer that's gone down year-on-year there, so everybody is selling more MIPS than they were this time last year, and we know it's still getting designed into lots of places and lots of that customer base. So, no sign of any sort of long term decay there at the moment where we're still getting decent volumes from there.
And then PowerVR, that 191 million number is pretty much where we expect it to be and you will understand given the way our months folds, that's very much in accrual. The big number in it is October's accrual which is one-third of the Q4 number which is obviously our biggest quarter in the calendar year for royalties.
And the royalty statements that have come in post the putting together of these numbers have suggested that 191 million is an under accrual, so we will – when we report in the second half get the upside from under accruing that 191 million in the first half. So when these numbers will put together the September royalty quarter statement were not all in. We've conservatively accrued and do not expect that number to be well north of 200 million more like 215 million to 216 million, when we get the final actual.
So, it'd down year-on-year in these numbers. We've accrued at what level we expect it to be and we've probably under accrued, but nevertheless it's on track and if we get some upside in the second half, that's all the good for that.
Ensigma slightly down on royalties, but up quite a lot on in the licensing line. We'll come to that on the next slide.
Costs, I spoke about is a £4 million FX number in that £52.3 million you see there. So, really like-for-like cost down about 10% year-on-year and that will run into the second half. The exchange rate at (1.25, 1.27) where we're at the moment, we'll have a small impact in the second half because of those dollar based cost, but it won't significant move the dial and we'd expect to be well within forecast there.
Just mentioning one or two of the other items further down the P&L, share based payments you can see quite a big jump there. What we've done in this financial year is the Imagination had a history of doing two share awards to staff a year. We've done one and we've done it in the first half of the year and the reason for that is retention, so we've put in a very competitive labor market, we've put retention in for the stuff that are there in the continuing operations and that's had an impact on the share-base payment. Obviously it's not cash, but it's increased the charge above where it would have been last year, because we've taken the full year in the first quarter – sorry in the first half.
And then the other thing says about net financing expense there, you can see a big jump year-on-year, almost all of that jump is exchange, so we have had to retranslate our dollar debt at (1.22) I think the October closing exchange rate was and that's at a £4.6 million impact on both on the value of the debt and on the net financing expense. So, you've got a big increase there, again not cash and if the exchange rate doesn't get any stronger, you won't see any further movement in that in the second half, but it's there in the numbers. Everything else I think as we would expect it to be.
Moving on to the divisional results, some of this I've covered, but PowerVR's licensing is about £5.1 million, £5.2 million for the half year and slightly down on last year, as we say in the statement that PowerVR licensing has been slower than expected and has probably moved to the second half, and that's shown in these numbers and it's probably why you've got to tick down year-on-year there and also the royalty volumes that we already spoke about, royalty rates are pretty much flat year-on-year, we're not seeing any reduction in those.
MIPS royalty rates are down slightly, that's a mix thing, probably down £0.005 on average from £0.05 to £0.045 that seems to be a seasonal thing. I'm still learning about some of these royalty trends. I've only been here just a year now and MIPS seems to go up and down like that if you look over the last three of four years, you get certain trends depending on which customer cells which chips in which quarters. But no underlying issue there we think as I MIPS already up.
MIPS licensing was flat period on period, but they did get a number of deals out there single-use licenses which should bring royalties, most of them came from existing customers, so we'd expect to see those coming through in royalties in the future, so there is some growth there. And then Ensigma's licensing was very strong, some large deals with big customers and you don't see by any means the full impact on revenue of those deals yet.
Some of them were signed in October and because of the nature and the size of them, we would – we will recognize the revenue on a percent complete basis, so we're not taking a big hit or a big advantage upfront, so there is more revenue to come in the second half and in FY 2018 from those Ensigma licenses as they go through their full licensing period.
And so Ensigma's jump in revenue should be maintained to a degree for the rest of the year. It still doesn't have a great deal of royalty in it and that royalty won't really start to come through until FY 2018, but it is a big improvement and it is now covering its engineering cost for the first time. If I look back at some of the Ensigma's numbers through history, it hasn't done that in the past and for the first half of the year, it's – and engineering cost number was £4.4 million and its revenue £4.4 million. So, it's on its way to profitability and Andrew will cover some of that in the later slides today.
Just a couple of things to say about the profitability of the businesses, what we've done here for the first time is we have fully allocated all the costs of the business. So, Imagination has always held things like sales and marketing, customer support costs in central cost centers and just have until actually the full year results never shown them split out at all, but we did show them as a large lump. We've taken that large lump now and allocated it properly to the businesses that it goes to. So, you can see the underlying profitability of the businesses.
So PowerVR is sort of a high 38% I think that is operating margin business. And if its license number comes back, and it will always has a second half royalty increase anyway, because it's the Christmas quarter of the biggest customer, and various other customers. So, it should be comfortably a 40% operating margin business and a good license year, we'll see it pushing to the high 40%.
MIPS is on its way to profitability, and we said at the previous results that we thought it would make in 12 to 18-months' time, if it has a good licensing in the second half, it can make profitability with all costs allocated So, this is not just on its engineering cost.
Ensigma's got a bit more of a way to go, it's covered engineering cost, but it's still obviously carrying a share of the other costs here. It's a big improvement on last year, but it's still got some way to go to get to breakeven and then profitability. And as I said, Andrew will cover that in the later slide.
Everything in that central costs line now is basically what it cost to be a PLC, so that's the Board of Directors, the advisors, some of which are in the room and it will stay at that kind of level going forward and certainly in PLCs I've worked in the past, that's a more normal set of cost to classify as central and we'll carry on doing that.
The final finance slide is the cash flow. I mentioned the large outflow from discontinued operations. That nets with a £9.5 million inflow from continuing operations to give you that £1 million cash generated by operations outflow there. The other couple of numbers to pick out are probably the CapEx and the disposal numbers, so CapEx as we said at the full year, down substantially, because we are not building buildings anymore. We were still building in the last financial year, the third building on the Kings Langley site, that's complete, completed in February.
So, the CapEx outflow you now see is probably a more usual one. We buy replacement IT kit and we file a lot of patents and we capitalize those patents, obviously because they've got a 20 year life. And we probably spend, £2 million to £2.5 million a year filing those patents and actually we've been filing a lot more in the last year, 18 months than we've ever done. So, we're still inventing plenty of stuff. And while we don't capitalize R&D cost per se, that is essentially as near as we get to capitalizing R&D, capitalizing the patent application process.
It takes through years, it costs a lot of money with external lawyers in various countries when you are filing China or in the U.S. and all of the other markets that we do business in. And so it's a big number, but it goes on the balance sheet for – it doesn't go on the balance sheet for 20 years. I have to say, we write them off over 10 years, not 20 years, but nevertheless it's a cash outflow every half year.
Then the disposal proceeds, that's largely what we got from Pure and you should see in the second half a large number when we sell the building that Pure sits in Kings Langley sale is going through and it will hopefully complete in the next few weeks.
Finally, on the net debt number there, most of that difference on October last year, it was £33 million in April, £35 million last October. Most of that jump up to nearly £41 million is a £5 million exchange loss where we've converted, so we're not really borrowing more money in any particular currency. And because we'll pay back in dollars, because we obviously get dollars in from our customers, it won't impact the cash, but in the translation of the balance sheet amount, you got a £5 million increase in the net debt there caused by exchange.
Then a little bit of an outflow from the April position of a couple of million because of these costs that we've been paying for the discontinued operations. Those cash outflows haven't quite ended yet. IMG Works hasn't gone. If you look in the detail of the segmental notes in the statement today, you will see that IMG Works lost about £4 million in the first half of the year, and that loss will run into the second half, although because of the way it's contract unwind, it will start to generate some cash in the second half of the year and then going into FY 2018. So, it's impact get smaller from now on, but nevertheless it's still burning cash at the moment and that's why we've got a net cash outflow in the first half of the year.
Final comment was about covenants, really a comment back to the full year when we had a covenant breach April, that's all been renegotiated as we said, we're comfortably within all the covenants that have now been set and don't see any issues going forward and there is still plenty of headroom in the facility. So, the cash position is more comfortable than it was a year ago.
That's all I've got. Back to Andrew.
Andrew Heath: Thank you Guy. So, let me talk about how we're doing in terms of execution strategy. So back in July when we were here five months ago, I think I said that we were close to completing the operational review that we kicked off, that's now done, well that was complete and I think I said in July, we were absolutely blessed by choices and I still think we are completely blessed by choices and I'll talk about those choices and which choices we are selecting later.
Within our core business of PowerVR, MIPS and Ensigma, we are very much sort of focusing those businesses that we decided they are the core of Imagination. So within those core businesses we were also being very clear about what we are going to do ourselves versus what we are going to partner and how we are going to focus those three activities, and particularly where our competitive position is strongest and we can differentiate ourselves, make sure we have leading disruptive technology that we can build strong defendable positions in the market we want to serve.
So maybe just sort of quickly sort of walk through, one of those differentiation of PowerVR is all about maximizing the strength of its leading architecture. We have the leading architecture in the industry still with PowerVR. We are going to continue to invest there. We are deeply embedded in our customers' ecosystems. We are going to maintain leadership and we are going to continue to invest in ray tracing. So for PowerVR, it's all about being a technology leader, staying out in front and continuing to support our customers.
In terms of MIPS, we have decided the MIPS will going to focus on the embedded process of space, and in particular those markets where we can use our unique technology advantages versus the competition, and particularly around multi-threading where we have a unique capability certainly compared to arm and we can offer multi-threading through all of our costs. We can also offer hardware virtualization through one of our MIPS core which provides a very strong security solution, which is going to be increasingly important both in terms of IOT, but also in terms of artificial intelligence, automated driving, et cetera, et cetera.
From an Ensigma perspective, what our customers really like about Ensigma is that we can offer end-to-end solutions that we can offer the analog the RF. We can offer the hardware design, the software stacks, certification against stand as a complete solution and that's why we seen in Ensigma licensing picking up in the first half of this year and we're confident, we'll see that continue.
To reiterate what I said back in July, we are seeking to address ecosystem weaknesses in MIPS and to help grow MIPS and Ensigma in terms of scale much quicker by looking at partnerships around the world to make sure that we can grow those businesses quicker.
As I said in my opening, within all this we are very much targeting profitable growth being disciplined about how we allocate capital and making sure that we focus on shareholder returns. But to be clear, the investment is underway, the strategy is being executed and we're moving forward.
Let me just quickly talk about sort of the growth markets that are out there that we decided we are going to serve, so the mobile, automotive, IOT, consumer networking, AR and VR, the operational review will spend a little bit of time looking at where we are going to play, how we are going to win and these are the six market segments that we are really focusing on in particular when there are others that we serve at least the ones that in particular we believe we can apply our technology and differentiate ourselves and provide strong solutions to customers.
So, let me just talk quickly about each one. So, firstly for mobile, our focus is certainly there in terms of graphics and multimedia with PowerVR certainly growing market share in mobile, particularly in the mid-tier markets, where the Series8XE has being seen to getting a lot of design wins and proven to be very successful. It's been the most successful product launch that we've had as a business so far.
And within mobile for MIPS, as we talked about mobile LTE baseband modems is where we're focusing the MIPS activity mobile. Again, we've getting some strong design-ins and we'll start to see that flow through in terms of royalty returns next year. So, for mobile, even the market is flattening off in terms of overall growth rates. We are looking to take share and overall growth losses in the market.
In automotive, you know automotive I think as you go out and pick up the newspaper every day or watch the news. Automotive is going through a huge amount of change, a huge upheaval, likes of Tesla, Uber, Lyft are really causing the incumbent manufactures I think differently about how they are going to compete in the market and they are accelerating that plans to change.
Every day, we see something new, I mean including sort of apples commentary over the weekend about them getting into automotive, into cars. I mean there is a consequence of all of that when you add all that together, we are now having conversations with OEMs and Tier1 suppliers in the automotive industry about them now considering designing their own SoCs, their own applications have made them more competitive and reduce time to market. So that the move to autonomous driving, more processes in cars, more security in cars, all of that work is certainly accelerating.
And if you just look at the amount of processing power that's going to be required in cars as you go forward. If you look at sort of, fundamental ADAS, fundamental ADAs requires about a 10X improvement in processing, you got a semi-autonomous is about 100X and fully autonomous driving is about 1000X increasing in process of performance. So that's going to provide demand for CPU we'll also shift quite a lot of the work on the GPU applications as well. So, automotive is definitely a growth vector for Imagination as we look forward.
Within that, we have a strong position already with PowerVR and automotive infotainment. That's going to go progressively higher end in terms of HD displays, 4K displays will leverage our proposition there to move into other applications in automotive and clearly MIPS also has a strong position with Mobileye, the leading vision camera supplier into automotive industry, which again we intend to leverage our position.
In IoT is all about security, I think we are all seeing that's becoming more and more and more of an issue. We've known about it for some time, but is certainly building a head of steam, our hardware visualization capabilities across our MIPS cores as I talked about is a really opportunity for us to establish ourselves in the IoT space, where today a lot of the ecosystem has yet to be established certainly in full, but we also offer hardware virtualization through our GPU cores.
So, as you look to things like autonomous driving, then we can apply that same technology in GPU and in fact we are doing so with Renesas, the latest R-Car chip that Renesas produced were automotive infotainment and their displays basically uses the hardware virtualization capability without our GPU and they can run three different operating systems on one chip because of that virtualization. So, even if it gets a tax, the infotainment system gets a tax, the displays and engine controls will still carrying on functioning. So that's a very strong solution and again an exciting opportunity for us.
In the consumer, we maintain a very strong presence in digital TV and set top boxes both with MIPS which typically have 40%, 50% market share in that area. But also with PowerVR and the Series8XE that we launched for mid-tier phones has been successful in mid-tier phones is also proving to be a great fit for digital TVs, we tuned the core, our high-end core, 8XE cores down, removing some of the overhead that's not necessary for mid-tier phones very much focused on fill rate, making sure the user experience gets a very rapid response in terms of fill rates on the phone. So, because it's focused on fill rate, it also work very well in the TV market, which is all about refresh rates on the TV. So that's proving to be very successful and there is another good growth vector for us.
In networking, again we have a strong position in networking, particularly with MIPS. As you see in the results, Guy has taken you through our actual royalties are up in MIPS and all of that comes from the networking area in terms of connected home routers and networking in general, and that's a home LAN for MIPS that we are going to continue to focus on and make sure we have the right products aligned to that market.
And then in terms of AR and VR, this is very active, there are all hype around, augmented reality and first reality? Yes. Is the hype likely to increase, I think it will, but it's definitely coming, and the conversations I am having with customers are all looking at AR and VR and how they can exploit that technology.
I think that will be certainly a growth vector for us over the next two to three years. And then as you look beyond sort of AR and VR and look at sort of artificial intelligence, then again that will be another growth vector for us maybe in sort of the next quarter five time horizon, again providing opportunities for all three of our businesses.
So let me just quickly then just walk you through where we are with each of the three core businesses in turn. As I said for PowerVR, our focus is on consolidating opposition and retaining leadership in the high-end mobile phones. We're also regaining share in the midrange mid-tier mobile phone markets, because the said opportunities also exist in automotive infotainment and digital TV, set top boxes and also in AR and VR for PowerVR. As well as exploiting our technology that we've been developing in terms of ray tracing which certainly in the high end AR/VR part of the markets will be I think truly disruptive as it will be in gaming as well.
So, one thing that I would like to be clear about is that, there has been quite a bit of speculation in some of the sort of industrial press around sort of how embedded is our technology in one of our leading customers. I just like to reaffirm our view is highly embedded. Their ecosystem has been built over many years off the back of a lot of our IP, and I'll just say the ecosystem they have established is high reliant on a number of the architectural aspects of our IP, which I am sure therefore will endure going forward.
As I said, the 8XE has been a very successful product launch. It's been actively designed in by number of licensees. And as a consequence we expect royalties and unit shipments to increase as we go into calendar year 2017. And we are seeing that being licensing in mobile, in digital TVs and in automotive. So it's getting a very broad pickup from a number of customers.
We've also seen that Mediatek have announced that our GPU architecture is in there Helio X30 product, and as we look forward we fully expect a number of the developments there to continue to grow, but also we'll start to sort of intersect some of what we're doing around our architecture and GPUs with things like Google's Daydream VR specification, untethered headsets which will give us further opportunities in terms of VR. But we do today have products shipping, IP shipping and products, it's in the Christmas markets in terms of VR headsets, 360 degree video headsets.
That's with some of our existing cores. Our newer design of cores are going to further enable and shift in performance and shift in quality. And if you look to where to have to really unleash the augmented reality and virtual reality, then we believe our architecture has the power to do that because the two key problems that need to be solved, I think to make virtual reality go as a mass consumer product is improve the performance at a very low power, so you can get untethered. So, in a high-end VR headset today, tether devices, I don't think many consumers are going to tolerate having that headset wired back to a box, it does limit the movements and restricts them.
But the other thing is the latency on the refresh rates. So as you render the image, you don't get a very fast response and you get a lag and that's typically what you read about if you looked any of the VR reports, where people get this motion sickness or they feel dizzy. And it's because the image isn't refreshing fast enough in our brain, it understand the image isn't refreshing fast enough and it causes the brain to feel to get out of sync, and therefore that creates the motion sickness and the dizziness and the nausea that goes with that.
I mean if you look at high-end devices within our existing architecture, but also ray tracing in particular, we can address both the power performance question and we can also address the latency question. So I think we have the technology that will enable VRs to become a mass adopted.
I talked about automotive infotainment and we have strong positions there today. We are licensed-in by Renesas and TI. They are very healthy relationships and we continue to see royalty rates grow in auto infotainment. And what I said in terms of Renesas R-Car product, you can just see the power of our architecture and how that can be used going forward into the future in terms of building opposition automotive and particularly ADAS assisted and automated driving.
In terms of ray tracing, I think we've announced an collaboration with OTOY. I think it's just a testament to the power that ray tracing has in its truly disruptive nature. I am excited by the opportunities as I said to get that into a sort of VR headsets, but I also think you'll be a much – you'll create much more immersive gaming experience. So we're talking to a number of players around the game console industry and how we can apply the ray tracing technology into that market.
And potentially, as we look forward and we look towards vision recognition, artificial intelligence, then we'll see more of the workload shifting from the CPU onto the GPU in that environment using the parallel compute power of the GPU. Again our existing architecture in combination with ray tracing I believe will be a great opportunities for us in sort of growth vector and maybe the sort of four to five year horizons subset.
And finally, we've been very clear, Guy mentioned this that we fully anticipate PowerVR to be at least a minimum 40% margin business, so certainly trading in a 40% range margin. That's very much our intention and to get back to those levels we have been there in the past to get back those levels as quickly as possible.
I think one of the final point on PowerVR before I move on, it's just about recruitments and employees. Again, there has been quite a bit of speculation in the press about us having a brain drain out of PowerVR going to our customers and others in the industry. It's fair to say that over the years we have seen continued movement of people from Imagination into some of our partners and other tech companies, but equally we are continuing recruiting from our partners and other tech companies and graduates and others as well.
So it is business as usual as far as I am concerned. It is a competitive labor market, but we're actively recruiting and we are growing our resources. So, we are well on top of recruitment and making sure we have the right people and right talent to grow PowerVR going forward.
Moving on to MIPS, as I said back in July, MIPS is being refocused on to the embedded processor space. This is where MIPS is strong today and our MIPS has a stronghold for many years. We do continue to get design-ins in digital TV, set top box, connected routers as I've said. We are making good progress on mobile with the LTE baseband modems and also in automotive, and as a consequence are getting this – the refocus of MIPS concentrating on this core embedded spaces. So, in my conversations with customers, customers in this area, and in particularly some of the changes that happened with the market dynamics over this current year, there is – I would say a greater desire to look at Imagination Technology, look at MIPS technology as a consequence of some of those shifts in the market, but also the fact that we are now focusing much more strongly and our product roadmap being aligned to customers in this space and making sure we have the right product at the right time to satisfy their needs.
And our differentiated technology certainly in terms of multi-threading, and that's the reason we got into mobile, baseband, modems, it's the reasons we won Mobileye and their EyeQ5 products, so that differentiated technology again is another touch point and why we believe MIPS can be successful in those markets spaces we've selected.
And from an IoT perspective for MIPS, the security solution is what's going to really enable MIPS to be successful in IoT built a strong ecosystem. The hardware virtualization that we have in all of our cores whether it's from a microcontroller, all the way through to a server class core, does allow a very strong security solution and we believe the stronger security solution to be enabled in the IoT space. And if you know, you could just look back over the last few months in terms of the number of attacks on major companies, including now sort of using surveillance camera, another small connected device using those IP addresses to attack businesses, security in this whole IoT space is going to become increasingly important.
But as I said earlier, we do recognize. We have some ecosystem challenges in certain parts of the market for MIPS. That means in some areas we excluded ourselves, so we are not going to trying to compete MIPS in mobile, apps, spec apps processor space, but it also means in some of these markets, we do recognize that we need to build our ecosystem. So we are investing that in ourselves, but also right to be talking to partners around the world to help address some of these ecosystem points and also to grow scale for MIPS.
And to be clear, Guy and I have set a very clear objective for MIPS and that is to be profitable in this coming next financial year FY 2018.
Moving on to Ensigma, and it's all about now, it's all about focus on IP licensing for connectivity, the complete end-to-end, WiFi and Bluetooth solutions that we offer with by Ensigma have resulted in that strong license deal. I am taking the first half, so it's up to 2.6 million, as half year over half year as Guy took us through the number. So we are talking about that strong license pipeline for Ensigma for some time. It's really good news to start seeing that flow into license deals, which will ultimately flow through into royalty shipments as well.
In terms of notable deals, what caused that uptick in licensing in the first half, where we had made deals with a mobile computing with some large industry players, also unmanned systems and also in IoT and one of the deals was with a significant Tier1 provider as well. So I think that's all are important validation of the strength of Ensigma's offering.
And I think another sort of usual touch point for you is that we are now starting to see some of that connectively IP, some previous licenses now being taped out. So, we will start to see product being ships with Ensigma connectivity IP which will flow into royalties in this next calendar year. Typically, if you look at Ensigma revenues from a royalty perspective, most of it's been around old TV radio business and that's why Ensigma still gets its royalties from. Going forward, we'll start now to see more of the connectivity IP coming through in the revenue line.
Again, in terms of strengthening Ensigma's position and growing scale faster, we are talking to partners around the world about how we can work with them to provide fully compressive solutions, so again doing a again doing a [indiscernible] internally, to certainly where we are going to invest in, where we're going to partner, but also look at partners around the world to help us grow scale quicker. And again we've set a clear objective for Ensigma and that is to be profitable by FY 2020.
So let me just summarize, you know having been CEO for nearly 12 months now for Imagination, I am really excited about the opportunities that are presented to us. We have very meaningful relevant technologies and that customers are interested in. I have had many, many customer meetings over the course of the last 10 months. You know they – we get great access to customers. We absolutely sort of punch above our weight. I mean there is one thing that really struck me since I've come in, and we – you know focused on attractive growth markets.
I think we've absolutely got the business back on track. The restructuring is almost complete. We are on track to get Imagination back on to a sound financial footing and we're very much on track now to go and execute what I believe is a really exciting strategy focused on those growth markets.
Our underlying trading remains in line with our expectations and we fully expect to continue to see an additional benefit from the strong U.S. dollar as we go through this year. And we focus on FY 2018 or the calendar year 2017, we're absolutely on course in mobile with the designed-ins that we've won in PowerVR and also in MIPS in terms of a GPU and mobile baseband and you know more generally, certainly my conversations with customers, you know there is a hunger for our technologies as I've said and are seeing lots of opportunities to grow business across all three of our activities.
We have a clear refreshed and compelling strategy that absolutely convinced we'll deliver profitable growth, and it's becoming increasingly apparent to me as I've got into Imagination, to all the business, gone around with customers that we absolutely recognize for having great people, great technology and also great potential, and clearly it's my job and Guy's job to make sure we'll deliver on that.
So thank you very much. That ends the formal presentation. We're very happy to take any questions. Nick?
Question-and-Answer Session
Q - Nick James
Morning. It's Nick James from Numis. Couple of questions. First was on Ensigma, in terms of the Tier 1 license win, could you help us in terms of what type of application that might be and the potential volumes of that win?
Andrew Heath
So, the area is typically in mobile computing. So, potentially it will be in the sort of tens of millions depending on the success of the customer in terms of translating that into product.
Nick James
Okay. And then in terms of the royalty rate that we should be expecting on these newer wins in Ensigma, both the ones you just alluded to, which were about to start shipping, and then that one as well?
Andrew Heath
Again they vary. It's always difficult to be precise. I mean we would – I would say sort of Ensigma sort of sits between sort of the sort of lower end of MIPS and the lower end of PowerVR. It's in that sort of range typically speaking.
Nick James
And the second one is just on automotive, I think last time we met, you kind of talked about – it was quite a competitive market. It was kind of difficult to judge how good the royalty rates were going to be in automotive. I assume to be a little bit more kind of confident and enthusiastic about in your comments today, so can you talk about that and just also put it in perspective of the competition in that market, see NVIDIA's done very well quite early and you are with Mobileye, but that's kind of a low-royalty rate, one as some of these other ones are kind of the second phase, whether they are really kind of credible competitors?
Andrew Heath
So, I think – is it going to be a competitive market, and I'm still concerned. It might be a bit of a blood bath, so I would say we are still being cautious. You know I wouldn't – we are not all in if you like, but we clearly have strong positions in our auto infotainment presence over many years and also through MIPS as well. The auto infotainment, we are certainly being pull through by our existing customers into do more assisted driving stuff and they are incorporating some of the safety critical functions into single-chips.
So, I was wondering whether some of the industry was going to go sort of separate tier for safety critical and separate SoCs for the rest, but fundamentally they realized they are going to have to build strong security and safety into a number of factors. So, even the infotainment system is going to end up with a safety level not as high as full automated activities, but – so we're having to build functional safety requirements, since what we're doing.
So, I think we will build up our position, the fully automated, you know and higher assisted levels and automation in cars, we are watching carefully. I do think there is opportunities for us, but it will be – I think we will take a cautious position, because I think we've got to make sure we understand the competitive pressures.
I think one of the advantages that we have is that we do have a low power architecture both in MIPS and in PowerVR, and the danger of using sort of large desktop like CPUs and GPUs is typically they have to be cooled and automotive guys really don't like active cooling, because if that fails, then potentially it's a failure point. So, having low power devices that consumer less energy, is not so much of an issue from consuming the energy perspective, because the car could generate enough electricity is the fact actually that you can operate it almost at lower power levels and therefore low heat levels. And you know if the car is sitting outside in the sunshine, being baked, it gets to sort of 40 to 60 plus degrees in certain climates. Actually when you get in the car, you can turn the ignition on, it will fire up rather than having to wait number of seconds before everything cool down, so this will lead to power up. So, there is all sorts of things to play in there that we've got to work our way through.
So, I'd say, I'm a bit more bullish than I was five months ago, yes, but we feel like it's a cautious approach.
Sandeep Deshpande
Hi Sandeep Deshpande, JPMorgan Cazenove. Couple of questions, I mean we can see your earnings coming through now through your restructuring, and interesting to hear your future thoughts on in terms of the growth. My question would be that, is the growth trajectory that you are planning in terms of the graphics and the end markets in MIPS that you are targeting forward looking enough given that the market focus in some of the markets where you've been in has moved to things like machine learning et cetera. So, are you involved on those markets, or is it that you are still looking at some of the backward looking markets which like in the smartphone market and in some of those niches which are actually not going to grow as much, but clearly you are going to grow in some areas like automotive et cetera?
And secondly, in the automotive market, following onto the question, I mean your customer Mobileye has moved on to license to work with Intel as well, and so does that mean that in the future you get dropped, or is that you are going to be a co-processor with Intel?
Andrew Heath
So, I'll just try and break your questions down. I think maybe the – the best way to answer is, if you look at PowerVR as an example, the sort of immediate term growth areas are very much regaining market share in mobile. As I've talked about, it's growing in automotive along with our infotainment customers, which continued to grow more process as VP is going to cards and graphics going to cars and then the Series 8XE wins in digital TV. So that is driving growth sort of in the here and now for PowerVR.
The next sort of growth vector, I'm actually convinced will be AR and VR, so that sort of maybe two to three years away and beyond that sort of artificial intelligence. I don't see us getting into machine learning. The high performance compute end using GPUs and NVidia has done a great job in terms of carving out that space. Could we scale up? You know could our architecture scale up today? We believe it could and we think we should invest in it and try and break into that market with all the barriers to entry, no, we don't.
So, what's the take is there then, would have to be licensed for them to go there, but that's a not a market, but we're – that sort of sever farms, deep learning, machine learning, piece of it we're not looking at. But we're focused on it in terms of artificial intelligence, is at the client end and being – you basically once that machine learning has been done, you need processes to end, go and apply it in products. That's where we will use our high-performance low-energy devices, GP-GPU compute to support that. So, those are sort of three phases of growth that we see for instance in PowerVR.
On terms of your question on Mobileye collaboration with Intel, as far as I'm concerned and I know, we have David Hound here who maybe looks at this in a bit more detail. He can answer the question if I don't to your satisfaction. But I don't see that as a threat to our relationship with Mobileye. The MIPS processor is low-power multi-threading capability high-performance capability that's embedded in EyeQ5, that's the differentiator.
Sandeep Deshpande
…You will be the first mobile GPU that conforming with OpenVX 1.1 which is the API for the smart vision systems of CNN and other learning vision system, so that I think, well ahead of the curve there?
Andrew Heath
Yes.
Janardan Menon
Hi it's Janardan Menon from Liberum. Two questions, one is, your competitor, your main competitor has been bought by SoftBank and seems to be reorienting itself a little bit more towards IoT. So what is the impact if that has had on your business; A, on terms of PowerVR, do you think that they are taking their foot off the pedal at all? Does it help open doors on the PowerVR side for you versus where you were six – one year ago?
Does it also make your efforts in IoT on MIPS a little bit more difficult, because here you have a huge beast with even more resources who is sort singularly going after that market. So, what is the puts and takes of that decisions at your major competitor?
Secondly on ray tracing, I was just wondering, you are saying two or three years, when do you – when would you expect to sign your first ray tracing license, is that possible in FY 2017 or is that really an FY 2018, FY 2019? And in your slide, you mentioned games consoles, auto et cetera, but you didn't mention smartphones, is that a market at all where you expect ray tracing to play and is that sort of later on or is that sort of going?
Andrew Heath
Maybe answer the rate tracer one and then come back to just sort of to summarize your first question. So, on rate tracing, certainly a personal objective of mine will be to get our first rate-tracing license in the next 12 to 18 months. I mean that's you know – we are actively discussion applications and how it can be used both in the game industry as I've said, but also potentially into professional work stations. So, then I think, you know the AR and VR piece of this is going to come out, really we're targeting the very high end to start with, because that's where it can be most effective. But certainly, given the number of discussions we have underway today, I'm hopeful that we'd get first license in the next 12 to 18 months. And what was the second part of the rate-tracing?
Janardan Menon
Is smartphones…
Andrew Heath
Not really. I think it's – you know I think we – once it gets into gaming, then on the software side to get written for it, then maybe the high-end, there may be some – some IP blocks or ray tracing that gets incorporated, but you know it's not – you are not going to see rate ray tracing for instance with certain in hardware accelerator IP it goes into boost the performance potentially. That's not an area. That's will be consequence of the other things we do rather than start there.
Then your first question, I forgot.
Janardan Menon
The arm…
Andrew Heath
The arm thing. So, I think – I'll tell you in general, you know the semiconductor industry continues to change at a rapid pace. I mean it's consolidating rapidly. The SoftBank acquisition was a major move. I agree with you, there are puts and takes. There are opportunities as well as threats amongst all of that. You know clearly the threat is that SoftBank will (parley) the investment in to something that they pay lot of money for. You know and go after certain parts of the market hard. But on the flip side, I think it's – you know many customers are – is it's crystalized their view that the industry does need to be more balanced and therefore that creates opportunities for us and clearly we're trying to use this opportunity to go and talk to customers about how those opportunities could be realized.
Vijay Anand
Thanks. It's Vijay Anand from Mirabaud. Couple of questions from me. Firstly, one PowerVR, it looks like the current revenue run rate is about £100 million and I think the business on current FX and FY 2015 made about £130 million, £135 million. So, the question is based on the current licensing pipeline and the royalty market share that you talked about, can PowerVR achieve those sort of revenue levels which you achieved just two years ago. That's the first question.
Secondly on MIPS and Ensigma, clearly both products, both divisions are heading the right direction. But there is still a significant difference in the margin profile versus PowerVR and it feels like in order to achieve some of the long-term opportunities, you still have to invest quite a lot in both these segments. So, can you talk about how you think about maximizing value creation from both these segments, and also whether it's possible for these two segments to achieve while not PowerVR like margins, but something which is more akin to a royalty IP business?
Andrew Heath
So, I mean I think you can see in the numbers we put up today that with some of the actions we've taken and by focusing those businesses, we've reduced some of the cost burnt in those businesses, which just helped their numbers as well as, from a MIPS perspective, continue growth in royalties and from Ensigma's select growth in licenses. So, there is a combination of focus in the businesses on key core end markets as well as some of the benefits we've seen on the top-line.
In terms of sort of margin expectations, do I see – we can guess MIPS and Ensigma are up to sort of PowerVR levels not any time soon. So the initial objective is to get profitable and then to move on from there and we should be able to get them into sort of territory that is more akin to our licensing business over time. And so, that's the objective. To do that we need partners as well to help pull us through to grow the scale quicker.
I mean what PowerVR has done is, what a good IP licensing businesses do, which is to get to scale and that's what we need to get both MIPS and Ensigma too, and so that's our focus in terms of our strategy to get them to scale as fast as we can. So that cover all the questions or was there another piece of it?
Vijay Anand
And the PowerVR revenue question?
Guy Millward
So, I've got the FY 2015 numbers to hand, but a chunk of that is license. This is you know historic [indiscernible] and some of that PowerVR license revenue in FY 2015 included IMG Works type licenses. So, IP going into System on a Chip designs that were being done in-house, that won't be reprobated going forward, And that business didn't make any money anyway. I mean you see in the profitability of it, so it's not a sensible thing to carry on doing that. The thing I don't have in my head at the moment is what the royalties were like in FY 2015 and how they would compare now on a like-for-like FX.
Right. Okay, well that would be higher than they are achieving at ten moment and there must be some volume difference that's causing that. There hasn't been a royalty rate difference, and I guess it may be the biggest customer numbers that are influencing it. The licensing will be the biggest outfit and we have to get back to a bigger license number every year for PowerVR to get the revenue back to where it was before. We're not there yet, but it's hopefully going in the right direction. When we get some of this pipeline in the second half…
Andrew Heath
Just to be clear on IMG Works, the IMG Works had some very quite lumpy and large license revenue, but typically had a very thin royalty tail. So, within our historical numbers you had some of these IMG Works licenses coming in which caused a one off boost, but it wasn't – you know because most of you would analyze the business, you would say, license coming, that's good, that's leads to long-term tail on royalties. For IMG Works that wasn't the case. So, it did sort of mask the reality of it, you know not deliberately, just in terms of how you wobble it forward. We know we are now very now very clean, pure play IP licensing business with that, any of that.
Doug Smith
Hi, it's Doug Smith from Agency Partners. A few questions, I'll split them up, so they don't get lost. First on PowerVR, I saw that the share of PowerVR to your major customer increased slightly. So, what's going on with the norm, like Apple business, why is it seaming to be so weak? Is it a technical issue, or marketing issues? Is the core simply too large or too expensive?
Andrew Heath
I'd say it's all down to – it's around timing. So, it's about time – is basically caused that got that is now being replaced by other things. PowerVR lost market share certainly in the mid-tier and that hurt progressive the launch of the Series 8XE in particular reverses that trend. And so we are seeing, we are still seeing some of those actually designed you know the products coming to end of life and then replacement products not having PowerVR architecture in, but we are now being designed into new products and we'll start to see those, start to flow through in terms of unit shipment and royalties in 2017 calendar year as we've been talking about. So, it's a timing.
Doug Smith
It's not a coaster or die-size issues?
Andrew Heath
No. I mean we've got to make sure we have competitive cores and die-sizes that meet you know specific requirements and we are actively investing in PowerVR to make sure that we have a comprehensive range of cores to satisfy the broadest range of customer requirements.
Doug Smith
Right. Okay. On the Mobileye business, the Mobileye like relationship, and just to clarify that. Is that just a CPU business?
Andrew Heath
It's CPU today.
Doug Smith
So there is no GPU at a moment?
Andrew Heath
Correct.
Doug Smith
Okay. And just to go back to this mentioned like AI, it's a kind of question like what is AI mean to you? You said it's not like a neural network training business, I guess is not inferencing as it, I mean what does it really mean it, it's not a machine learning business I guess?
Andrew Heath
Yes. So we are not doing deep learning piece. That goes of server farms way, they show billion picture of cats to work out, what a cat looks like. That's not what we're – we're not using our GPU architecture is not been used for that. What it will be using though is that in terms of using own network accelerates in the GPU under GP-GPU for general purpose computing using the GPU, using its parallel processor. So, in the client device, we need a client device to determine, I need to understand why my cat is, then you can put our GPU in that device and it's already been trained to identify, but all the algorithms run on our device in the client.
Doug Smith
So, is that what they call inferencing, is that an inferencing application?
Andrew Heath
I don't know. I am not sure.
Doug Smith
Okay. And just a follow-up. Have you had any request for doing single precision or half precision on your GPUs?
Andrew Heath
Yes. So today we've got demonstrations where we're using our GPU to do object recognition in particular. And again, basically what – most of what we're doing is, what we call image signal processing and hardware accelerators, new network hardware accelerators. And so, we build that into our GPU offering, so once the algorithms all being trained up, you are applied into that, to put in say client devices side.
Doug Smith
Okay, great. Thanks.
Andrew Heath
Yes
Achal Sultania
Thanks. It's Achal from Credit Suisse. Maybe a question on your revenue from your largest customer, if I were to look at the correlation between units like your smartphone and tablet units are your largest customer, there seems to be, the correlation between your revenues from then and what they are shipping into the market seems to be moving away. So, basically what I am trying to imply is that, you're revenues are growing at a much slower pace than what the units at your major customers are growing for the last three and five years. So just trying to understand is that more to do with the licensing part of revenues within that or is your royalty rate coming down slightly in what you used to saw, three, four years back? Just trying that, it doesn't make sense what exactly is happening from numbers perspective, so any color on that would be helpful.
Andrew Heath
I think we've been clear that and we're clear and we can't talk about anything commercial sort of confidential. The best way to say is that, the last agreement that was signed with the customer basically had various steps in it over time in terms of the commercials, all that is built into our forecasting guidance going forward.
Adam Bowler
Hi. It's Adam Bowler from Deutsche Bank. On the MIPS side, I think related to the fact that the royalty per chip has come down and again. And I am wondered you kind of said it was seasonality related, I wondered where we could expect that to normalize? And then secondly, more strategic one, there is been a lot of oppositions towards Chinese investments in U.S. tech companies particularly. No wonder whether any of your major customers are express any concerns about your Chinese shareholder?
Andrew Heath
Okay. I'll answer the second one. You answer the first one. So, the simple answer is no. We had no concerns raised regarding our shareholder.
Guy Millward
They have to announce when they go over 3% and they announced when they went under 3%, but they're actually off the register entirely if you looked at it now. So that concern has gone away, I think.
On the MIPS one, yeah, what I was trying to get out there was, it was about £0.046 average royalty rate for the first half for the year, and the average royalty rate for the second half of last year was £0.05. So it has gone down. The first half of the previous year was 4.6%. So, I'm saying, I don't know if there is any seasonality in the type of devices being sold, that pushed that royalty rate up in the second half of the year. But I've seen that in two to three years off the royalty rates differing between first half and second half. But we don't see any underlying reason why it's going down and the volumes are up so, so it's difficult one to answer to be honest. You are aggregating 50 or 60 customers over some very different arrangements. Some of which are very old on therefore it got some very different pricing in them. And it is difficult to say exactly why the averages move like that, but there is no one reason why it's step down.
Adam Bowler
Can you give us any idea on where it might kind of trends or normalize going forward?
Guy Millward
The bigger customers' have royalty rates that are higher than the average. So, they are the ones that are growing and suggest that it might hold up going forward. Other than that, it is pretty much going to stay where it is. I think it's good as I can say at the moment.
Andrew Heath
Question in the back and then maybe we need to wrap up.
Unidentified Analyst
Just a couple of clarification ones from me if I could. The £5.1 million licensing far PowerVR, can you tell me if any of that come from your customers in the first half of this year? Then just circling back on earlier question on Ensigma, the tier1 licensing with the mobile computing play, is that a competitive win. Did you just place another third party IP provider or is that a new win altogether for the market.
Next to that, 8XE, largest, do these deals, do they work on a par silicon rate. I mean you're doing this on the dollar price on the silicon or is there anything capped in those deals that you can talk to? Thank.
Andrew Heath
I'll do the last one first. I mean at the very high end in sort of our – sort of in the mobile space, it's all about sweating, how much performance can you sweat out in silicon, because there is only so much silicon you can physically put in the phone, you just need to sort of maximize it. So, typically at the high-end, it has lots of functionality, lots of overhead in it, large silicon area, but we absolutely then sweat the performance out of it and that's where we've been historically successful.
In the mid-tier market it shifts from sort of an overall parent performance to very much a performance of unit cost. So I got to hit this performance point, but I'm only prepared to spend this and therefore you've got to offer a competitive performance at cost point. So there we strip out lot of the overhead and we tune the cores very much to meet street points in the – within the product range for our customers. And then, I've forgotten the first and second question.
Guy Millward
On the first question, the answer is yes, there were new customers in the PowerVR licensing in the first half of the year. The 150 new customers it was in low-single digit, but there were new customers.
Unidentified Analyst
[Question Inaudible]
Andrew Heath
Well, I'll say it's – I'm not sure how to differentiate it, right. I mean in the Ensigma space, connectivity is a very large market. A lot of it is taken up by dedicated WiFi and Bluetooth or combo chips and go to a very large players and buy them from. So if you don't want to do that and you want to lower cost solution and you want to integrate the IP onto an SoC, then you ought to have an in-house team or you're going to license it. We certainly have the most comprehensive WiFi offering, Bluetooth is a bit more competitive and then Bluetooth we are focused on sort of more on Bluetooth classic. And then some of other applications we are partner with where the people deliver competitive solution across that I'd say.
Unidentified Analyst
[Indiscernible]
Andrew Heath
Correct. Yeah, they are using our IP. That was your question.
Andrew Heath
I am conscious, that's it's now 22:11. Is there any last earning question? I'll be and Guy, around for a little bit longer, after I'll be happy to take questions over a cup of coffee or anything. If not, thank you very much for coming. It's good to see you all again and look forward to seeing you again in the near future.