Very good. Good afternoon. Welcome to this Asetek Q4 2018, and more importantly, the Capital Markets Update of Asetek. We are a good group of people here in the room, some investors, bankers, media, employees, myself, a couple of more of my colleagues here, and then we have, of course, whoever is following us via the web. Welcome to you all. Thank you very much for taking time out of your calendars to listen to us talk about our favorite topic, Asetek. A few practicalities before we go into the actual presentations. One is that you may hear a little bit of sound. We are in the middle of our factory here in Aalborg, Denmark. If you hear a little bit of sound from the factory, please bear with us. It is your money at work, you could say.
Furthermore, I have been equipped with this magical device, which through a combination of engineering and magic, will allow you guys who are following us via the web to ask questions that will come up here. I will try and read them out as we go. We do have a Q&A session at the end, but if there is something really important, by all means, bring it up. That goes for you here in the room, too, of course. We have a microphone.
Please wait with your question until I have the microphone at you, because otherwise the ones listening via the web cannot follow. If you are following via the web and you are not online, meaning that you are looking at it at a later point in time, please write me an email at investor.relations@asetek.com and I will make sure to reply to your questions as soon as possible. Another magical device.
If I hit the green button, I will change the slide. The agenda today is that we have had a very light lunch. We then go to the first presentation by André Sloth Eriksen , our Chief Executive Officer and Founder. He is here. He is coming up in a minute. He is going to be talking about building a gaming and enthusiast brand. We then go over to John Hamill, our Chief Operating Officer, who is also here, and he will be talking about how we are driven by the end user experience. John and André will do a song and dance together and talk about maintaining position in the data center market. Then I will finalize this section here, session here today, talking about financials with the headline of solid financial platform for long-time growth. Then we have a sum-up by André and a Q&A.
By all means, if there is an important question, bring it up as we go. With that, Mr. Eriksen, Chief Executive Officer and Founder, this is your thingy.
Thank you. Good afternoon to you guys here and the guys online as well. I've been looking forward for, I think six or seven months to this day because I've been so enthusiastic to tell about what I'm going to tell you now. I hope you think it's as great as I do. What I'm going to talk a little bit about is building a gaming and enthusiast brand, or in other words, we are actually going back a little bit where we started, 20 years ago this year. What we're going to talk about is basically a new branding strategy. You could ask a fair question and ask if we ever had a branding strategy. Now we do, at least, and that's what I'm going to tell you about.
I'm sure at least those of you in the room have already seen our new logo. What does a new logo mean? Does it mean more revenue? No, not necessarily in the short- term, for us, this new logo means, let's say a kick start. We are changing our logo to put down a milestone and say, from today on, it's a new Asetek. That's both to you guys, it's for our customers, and as importantly, it's also for our employees. I'm not going to entertain you with everything our new logo stands for, but I think you can get a flavor for it by looking at it. We are proud of it, at least. Why am I here to talk about branding in the first place?
That's because I woke up literally six, seven months back and thought a little bit about our position in the whole ecosystem, being in gaming, being in data center, no matter what. What I found out was that we are doing solid products and solid solutions for our customers. At the end of the day, it's our customer's brand that is promoted. Fewer and fewer people actually know who Asetek is, so they may go and buy a liquid cooler from NZXT as we have over here. Unless you are a hardcore enthusiast or have the same age as myself, you don't necessarily know who's behind it. I think that's important. I think it's important because we have the IP situation. I think it's important for people to know what they're actually getting and what they are buying, and that they get the real stuff.
What we decided to do, I literally gathered the entire management team and we spent two full days together here talking about all these issues, and the conclusion we have come to is that we will put our brand forward. Of course, we cannot turn our customers into Asetek resellers. We have to do it in a way where it's a brand behind the brand strategy. It could be dual branding. What it also means is we may actually go and launch branded products ourselves that we may sell directly to end users.
We may sell it to our current customers, but with a buy that says, "If you want to sell it has to carry the Asetek brand, and if you're not willing to do that, you cannot buy it." Here I'm talking about high-end products, like more where I founded the company 20 years ago. In other words, what we want to achieve is to make sure that when people buy an integrated liquid cooling solution, if they do not buy one of our solutions, then at least they know that they are not buying one of our solutions. Whereas of today, that's not necessarily transparent. I think for the first time in the company's history, we have a real marketing department. In the beginning years, the marketing department was obviously me. Other than that, it has mainly been product marketing.
The reason it's been product marketing, because it's always been in dialogue with our customers and say, "DM is the customer, what would you like us to build for you?" The other way around, we have developed some products that we thought would be interesting, and then we have sold it to our customers, and then they have rebranded it, so to speak. Right now, we actually have hired PC enthusiasts, branding experts, et c, and it sounds crazy that we have the largest marketing department ever, but please bear in mind, we are still only 100 people in the company, so it's not like we have 67 people of marketing employees. Nevertheless, it's a shift in the company, and I guarantee you are going to feel a difference and see a difference in our outbound marketing activities.
One of the first initiatives, and to tease you guys a little bit, you probably saw the woman installing the sign over there that says, "Asetek eSports Academy." I made sure none of you actually walked in there. You can see it afterwards because what we have in there is a full-featured gaming room set up. Right now there's 10 Alienware PCs in there. We have state-of-the-art hardware and surroundings in there. This academy, that's what you see in there when the presentation is over. This academy is really about engaging with the end-user community. First of all, if you live in the area or if you live in Denmark or outside Denmark, if you're interested in esports and you're interested in having, let's say, elite training facilities, you can come and use the Asetek eSports Academy for free.
That's a good marketing tool for us. It's a good way for us to engage with the community, get feedback, what works, what does not work. The eSports Academy is just a small piece in a bigger plan. We are introducing, as you can see on our new website, another brand called CoolNation. CoolNation will be an umbrella for several activities. For example, a global esports tournament. There will be global overclocking contests where, for example, the finalists of the esports tournament will go here and battle out the last round, and there will be streaming from this academy. So it's a great marketing tool for us that we have built. It's an investment for sure, but we have some good partners, Alienware and Razer, who has pitched in as well. It's really nice in there.
If we look at the adoption of our solutions over time, you can see that five years ago, no, seven years ago, eight, seven, eight years ago, we were roughly at 1 million units, and today we have passed the 6 million unit mark. When we claim we are the market leader, there's actually some substance behind it. As a part of this whole rebranding exercise, we obviously looked a lot what is our strengths, what's our weaknesses. One of the things that we found is that both from end users but also from asking our customers, what we found was that we deliver the best performance, the best quality, and the best reliability. What you do not see up there is we deliver the lowest prices, and that is our weakness if you want, but we think this is the way forward.
We are not actually going to change a lot about how we build products and how we sell or market our products because these are our core strengths. This is essentially saying the same as I just showed you in terms of volume, that, of course, the revenue has followed the volume. Innovation has been a core tenet for us, and it still is. If we look back to the early days of 2000 with the VapoChill product, the HP Blackbird, our first OEM design win ever, our first integrated coolers here, the Corsair H50, our RackCDU, our In-Rack CDU. Of course, I can derive a lot from these slides. One of the things is this VapoChill product.
I think it's fair to say that when we attend a trade show, even now in 2019, some 22 years since I launched the first VapoChill ever, we still have people saying, "Why are you not selling this product anymore? Why can we not buy it anymore?" Without revealing anything, because the decision is not made, this is obviously some of the stuff that we're looking at. Should we bring that back alive? That was just an example. We are obviously going to talk more about the data center business in a few minutes, but what you can also see from this slide, of course, it's not very granular, but what you can also see is that from 2009 and onwards, we have not focused that much on the gaming business.
Independent of the data center market, I believe that's a mistake, and that's what we're going to fix now. Solving thermal challenges and the more complex, the better. What does that mean? Everybody and his dog can make an air cooler. It's a number of fins, then you have an airstream flowing by it. That's not our business, because we cannot make money there. We are in the wrong part of the world. We are way too many employees for making money on stuff like that. What we're really looking at, and will continue looking at, is the innovation part. Where can we make cooling solutions for the markets we operate in even better than they are today? It's all about driving preference for our technology, of course.
We want the gamer, we want the enthusiast that enters the store to say, "I want an Asetek product." That's the whole idea about this. As we have the leadership, as we have DNA that says innovation and performance, et c, we are doubling down our effort on the gaming segment and the enthusiast segment. This slide is going to talk a little bit about that. We have gotten rid of our desktop segment. In all honesty, it'll probably take me a while before I remember myself, but there's some thinking behind it. One of the reasons why we're getting rid of desktop is that everybody who thinks about a desktop PC thinks about something ancient and dead. Desktop PC is a dinosaur in everybody's mind. Instead of calling it desktop, we have renamed it to gaming and enthusiast. What does that mean?
Because it's still desktop PCs. Yes, it's still a desktop PC, but it means that we focus on the part of the desktop market where people are building high-end PCs, where people are gaming. We think it's an even crisper way to state what it is we are doing. Data center, there's no new sexy names for that. It's still a data center, I think it's pretty precise what it is. As most of you know already, 95% of our revenue comes from the gaming and enthusiast side, the remaining 5% from the data center. I believe, we believe, management believes that we are looking at two large and long-term growing markets. We do not believe esports is over tomorrow. We actually believe this is something that has come to stay. We are already supplying global brands.
There's obviously more, we have a good part of it, we have market-leading solutions. We have our IP platform, please note when I say IP platform, I don't only mean patents. We also have patents. We have our products, we have our high volume manufacturing, we have our quality, we have our hub infrastructure, now we can also add brand. It's not going to change with one Capital Markets Update. That's why we hired the marketing department. I think when we meet next time, I think we will have a solid brand. It's not like our brand is brand new. There's just a lot of people who need to be reminded about who we are. Talking about the IP and talking about the platform, I believe we have a pretty good setup for achieving what we want to achieve in the longer- term.
We have our Silicon Valley offices. I don't think it's necessary to explain why. Why do we have an office in Texas, John?
Because I live there.
That's because John lives there. There's one more reason. That's because Dell is there, of course, and we have HP in Houston. Here in Aalborg, of course, we have some manufacturing. We have most of our R&D. Munich is also kind of a tech center in Europe. We actually also have someone, I'll get back to that later on, that will be working out of Brussels, closer to the EU. In China, we have, of course, our entire supply chain, but we're also beefing up more and more engineering out there. Obviously you also think, oh, that's cheap labor. No, that's not the point. The point is that more and more of our big customers are actually placing their R&D departments out in Asia. Therefore, we also need to be there. That's why we are in Taiwan and in Taipei also.
I think I'll take questions about what I just talked about now, just to structure it a little bit, because now we are shifting gear after this. Any questions or comments to the branding story I just told you? Yes.
What kind of cost do you expect to incur with this new branding strategy? The academy and so on? It's not completely for free, I guess.
Anders from SEB asked what the cost was. I think we will dive more into the cost details, Anders. I know Peter will do that later. I think I can reveal a little bit, and that is the investment in R&D on the data center side. All of you in here in this room, I guess, knows that we are trying to dial that down a little bit and keep that on a constant level. Seen from an outside perspective, you will not notice the cost of what we are doing here. It's not like our operating cost will skyrocket. Just to put things into perspective, building this eSports Academy, we are perhaps talking $200,000.
It's peanuts in the major scheme of things, it's not like we are initiating a huge investment initiative that we hope then to see return on investment 10 years down the line. Is that answer enough for now?
Last question from ABG. Will there be a wave of new product launches? Without diving too much into the guidance.
Yes. One thing is, of course, the social media side and the academy side and the gaming side, the whole idea is to sell more products, introduce more products, introduce more exciting products. Yes, of course, that's a longer-term perspective. I think even if we put all our best efforts in right now, the first product launch that you would see would be a year from now. It's just the time that it takes. From the branding perspective, getting new products out with our brand name on, that's absolutely something we're already working on. Was that the answer? Okay.
Any feedback from clients?
You guys are pretty much the first ones to know. There are more of our customers, our direct customers. When you say clients, let me interpret that as our direct customers, yeah. Most of them are actually pretty excited because most of our customers, they only sell products from Asetek, meaning that they also see the value. They also want to put forward our brand. I think the best example, although it's old now, is Intel, that if you go to the street and buy an Intel integrated liquid cooler, it actually says, "Cooled by Asetek." It was Intel approaching us asking if they could do it. Then, of course, there may be customers who would want and like a financial incentive to like our new strategy.
That's not necessarily bad because, just a made-up example, if we give people $1 in discount for carrying our brand, as long as that dollar is then spent on marketing our solutions, then it's a win-win. Can you pitch in, John? Do you agree it's been well-received?
Oh, yes. Absolutely. Customers have been promoting or at least requesting that type of support for some time. As André says, if you're only using Asetek, it's very helpful for them that they can leverage any marketing activity that we are driving or driving with them.
Good.
Good.
All right. We will be able to ask questions later on. I would like to ask one question myself, just to get it out of the way, and that is: Is this something we are doing to put out a smoke screen over the data center business? No, of course it's not. Number one, we have focused very hard on the data center business for five, six years now. We are not going to throw that away. Number two, we have focused so hard on the data center business, we have not had the efforts to do what we are doing here. Still, we have been growing pretty good for five years on the gaming side of the business. Wonder what will happen if we actually start to focus on it.
As you will see later on, all of these slides will offer you full transparency, and there's no hidden agendas here. This branding strategy is not something we made up two weeks ago and say, "How do we save the Capital Markets Day?" Just to be clear on that. All right, John.
All righty. Just for those of you that don't know me, my name is John Hamill. I am the Chief Operating Officer at Asetek. Just about to celebrate my 10th anniversary at the company. Just thought I'd mention that, André. Today, I am going to be talking about our newborn market segment, the gaming and enthusiast segment. Of course, you all formally know it as desktop. I'm going to add a little color to that market. We'll start right at the beginning with a reminder of where the revenue comes from. There's ostensibly two groups of customers here. We have OEM customers, and we have what we call internally DIY customers, do it yourself customers. OEM customers include guys like Dell Alienware, Hewlett-Packard, or HP, if you prefer. On the DIY side, it's guys like Corsair, NZXT, all names you're familiar with.
The OEM customers, they tend to take our products and integrate them into complete systems. There's actually an example of one just here. This is a Dell system, actually, there's a lot more of them next door. They integrate Asetek liquid coolers into the systems, and they sell them on to end users. Our DIY customers tend to sell our products as standalone units to enthusiasts, and those enthusiasts are building their own systems. As you can see, there's quite a dramatic difference in the shares. OEM, around about 20% of our business. DIY, near around 80%. We shipped almost exactly 1.1 million units this past year. That says we're in excess of 200,000 units into the OEM business, just under 900,000 units on the DIY side of the business. We've developed this very, very simple graphic here just to help explain our view of the market.
Our market is described here as hardware enthusiasts, as you can see, most hardware enthusiasts are gamers. That means there's a lot of strong parallels between the hardware enthusiast market that we target and the gaming world. Maybe that's just as well, because there's not a lot of data specific to the hardware enthusiast market, our target market. We do find that we have to leverage a lot of information from the gaming world. A word of caution here, there's not a one-for-one analogy, and sometimes I do shiver when I see some of the hype around the esports market in particular. Okay. Just to touch on something that we've talked about over the last couple of CMUs. Previously, I've described an environment where gamers have this insatiable appetite for what I describe as the immersive experience.
How real can it be? That's not the only reason people buy liquid coolers. Any hardware enthusiast will tell you liquid coolers are far more efficient than air coolers, that's really important. If you're trying to overclock a CPU or even a GPU, a graphics processing unit, that's critical. If you're trying to build a lower noise computer, that's critical. Liquid coolers have actually seen a lot of innovation in the area of industrial design and aesthetics, that turns out to be really important to a hardware enthusiast. Why? They care about what the inside looks like. They customize the interiors of their personal computers. Lots of reasons beyond that immersive gaming experience for buying liquid coolers. We believe the fundamentals of our market are strong. That's based on information we have from the gaming world. All the indicators and metrics are extremely positive.
There's three separate charts here. Starting on the left, we have a chart that describes revenue from video games, from advertising within video games, and even from esports. We see some very strong growth predicted through 2022. I don't think anyone in the audience is guilty of living under a rock recently, you won't be surprised to see the huge growth rates predicted for esports revenue. Just to clarify what we mean by esports revenue, we're talking about tickets and merchandising from events. We're talking about media and the sale of that media. We're talking about sponsorships and advertising revenues. Very, very strong growths there. The last statistic, the chart on the right describes the growth in virtual reality headsets. Going back to that immersive experience, that was one of the key drivers we talked about in previous years. Absolutely formidable growth rates forecast there.
I can tell you if this is anywhere near right, this basically predicts if you're playing games, you're going to be wearing a headset regardless of whether it's a PC or a games console. Some data closer to home. We have some data here describing trends in the high-end gaming PC world from OEMs. This data is specific to the OEM segment we service. Think again about Dell, Alienware, Hewlett-Packard, and the like. We focus on high-end because we know in high-end systems, liquid cooling is common. As you move down through the price stack, liquid cooling can become an option or even not listed at all for cost reasons. We tend to focus our attention on the high end. As you can see from the graphic, we see growth of around 3% per annum.
What's really interesting is that the growth is dominated in the higher price points of these already high price points. This high end, we're talking above $1,800. As you can see here, price points above $3,000, we're seeing pretty dramatic growth. Same for above two, below three. What's interesting from our perspective, as I mentioned earlier on, that we sell just over 200,000 units into this space. We can do some CEO level math. Apologies, André. We believe the majority of our sales are into this segment, at least 2/3. If you take 2/3 or more of the 220,000 that sells into that segment, it tells me our market share is in the 10%-15% range, depending on how much more than 2/3 are selling into this segment. Just to make it clear, the balance is more than likely Air coolers.
Air coolers will dominate the balance there. We got some new data from Jon Peddie Research to refresh a view that we shared last year. Jon Peddie tells us the high-end gaming population is now about 26.2 million. That's 26.2 million individuals who own computers costing in excess of $1,800. To keep this analysis relevant, we have to take out notebooks, but that still leaves us with a population of 20 million users, each with a system, a desktop system, valued in excess of $1,800. We know these users refresh their systems every 3-3.5 years. That tells us annually we have an opportunity around 5.5 million-6 million. Again, I know that I sold 1.1 million units across the entire business, and I'm guessing the majority were into this space, at least 2/3.
That tells me that my market share is between 13%-19%, depending on how much more than 67% were sold into this segment. All right. It gets really interesting here. We're going to talk about what happened last year and start to look forward to next year. Last year was very unusual, guys. Very unusual. Ordinarily, I would have told you that Q2 would have been a low point of the year as it is every year. Q2 turned out not only to be our highest quarter of the year, it's our highest quarter ever. That was really unusual. As we moved into the second half of the year, we faced or encountered very strong headwinds, and those headwinds included macroeconomic issues as well as industry issues. From a macroeconomic perspective, we of course had the specter of the U.S.-China trade war.
Customers were also concerned by Brexit, the economic situation in Italy, so on and so forth. We had industry issues. We had the delay in NVIDIA's Turing architecture, an architecture that subsequently suffered very poor market acceptance. We had Intel and product shortages unimaginable, and we had the whole cryptocurrency issue, and I have had customers tell me they've been hurt by cryptocurrency, believe it or not. Looking forward, I can tell you those headwinds prevail, and it will impact Q1. However, I think we'll be through the worst of it in the first half of the year, and I'm predicting a strong second half. Indeed, as we look at the company's outlook, so the entire company, we're reasonably optimistic that we're going to see growth in the 0%-10% range. Okay?
We've been reporting on how we've been able to diversify our revenue base. We made significant progress in 2017. We looked at 2018, we were surprised to see we hadn't quite made as much progress. When we looked at the second half of 2018, we saw a much better picture than something that was broadly in line with our own expectations. What I will say is that that second half picture includes a new player in our top five, a company called ASUS, or you may also hear them referred to as ASUS ROG, referring to their Republic of Gamers brand. I believe we continue to make very good progress towards our efforts in diversifying our revenue base. To wrap up this particular section, be in no doubt we intend to dominate this space.
To go back to what André was talking to earlier on, we're going to do it by building our brand, by innovating, and yes, the goal is to bring our own branded high-end products to market. With that, I'll conclude the monologue here and open it up to questions. Do you have a microphone? Cheers, Peter.
Any questions for John at this point? Mr. Knudsen.
Hi, it's Anders again. Of course, I can't help asking about the elephant in the room, the decline in Q1. How big is that gonna be? Also, I know you don't have much visibility, when you say strong growth in second half, what kind of numbers are you thinking about? Because obviously Q2 was a very tough comp given it was such a strong quarter last year.
Based on our models, whatever softness we see in Q1, we think we can offset in Q2. Beg your pardon, second half, excuse me, Anders. Second half of the year. We are fairly bullish based on how we model the business based on what we can see to date that the softness in particular in this first quarter will be offset.
Is my mic on?
All right. Let me just add to that, Anders, because there are two things worth remembering about our business. Number one, we'll always have one bad quarter or low quarter, but we cannot say which one. That's the first thing to remember, and it's not even a joke. You can go back eight years and look. Number two, I know that you guys like Excel spreadsheets, and that's your job, but I warn you, as I've done before, do not read anything into whether one quarter is lower or higher than it was the last year or the year before, because it doesn't say anything about the future, in this case. In terms of Q1, it's no secret it's going to be low.
We don't know yet how low, because our visibility is short, but in my mind, we have to go back a few years to see a Q1 this low. What you also know at this time of year when we're doing this guidance, that we are trying to be conservative and trustworthy. If we just forget the data center for a second, our actually running business, we have been spot on or ahead for what we have said in 20 quarters now for a row, and we would like to maintain that record. When we say this full year guidance, that is actually what we conservatively believe in, independent of what Q1 or Q2 may or may not show. I think just to add a little bit more color to it.
Just another question on the guidance. Is there any GPU penetration embedded in the guidance? Or do you see any progress in the GPU penetration?
It's modest. It continues to be modest, single-digit percentage of overall sales.
I would like to add some color on that, because that's short-term. That was a short-term question and a short-term answer.
One, two years out.
Yes, because what I would like to add as a part of what I talked about, what I can say for sure is that GPU is a big part of those plans.
Any progress on any-
Well-
It's a diamond, right?
I cannot speak on behalf of my customers, because I'm not allowed to do that. Let me put it like this. We have some very exciting GPU stuff going on with a customer that says, "Yes, I am going to buy this if you can do what you say you can do." That's as specific as I can be.
One more question.
That's basically meaning that the market size will double the one you showed on the Slide 2 before, right? That's one unit per PC.
In theory, yes.
Now we have the Excel exercise again. Let me just back off for one second. What's preventing GPU adoption to be really big right now, I believe, is the fact that you have two hoses. We have a system over there for those of you in the room who can see it. You have two hoses coming out of the graphics card, and then you have a radiator and a fan. That is prohibitive for some. It's really difficult to go and buy a graphics card and then in advance, make sure you have the space in your computer case for the radiator. The trick is, can you do a liquid cooling system that's more efficient and more low noise than an air cooler, but within the same factor, as form factor as a normal graphics card? That's the big question.
If you can, then yes, then the market would potentially double. That's, of course, what we're working on. I don't think that's a big secret. We are not 100% there yet, but we believe we can get there.
Very good. There's a couple of questions down here. Please state your name, sir.
Thank you. My name, Poul Ernst Jessen, Danske Bank. I have two questions, and the first is also relating to growth. If we look at the gaming segment and next to what you have up here on the screen, the OEM segment and the DIY segment, where do you primarily see the growth going forward?
I don't think the mix is going to change too much. It really hasn't changed in recent years. I'm not expecting any high drama. There was a period in time where DIY was growing faster than OEM, but I'm not sure that's the case any longer. My guess is, and it is a guess, it's likely to stay similar.
Yeah, okay. Perfect. Secondly, it's probably concerning your new strategy more, but you talk about making your own products and releasing those to the market. Does that also mean you're actually going to compete directly against your OEMs?
No, that was a great question because there I was not clear. We are not going to compete with our customers, for sure. I believe fundamentally that's a bad idea. I can give you an example. What we sell today, as you know, is this integrated loop liquid cooling. The dream scenario for me would be to have a Corsair or an NZXT or ASUS branding their own product and then say, build on Asetek technology, for example. There is a market above that's what I meant by getting back to our roots. There is a market for people who are building liquid cooling components. They look very high end. It's like Hi-Fi in the Hi-Fi industry, right? Look very high end, they feel very high end, but the end user have to mess around with liquid on their own.
I do not believe this is going to be our next greatest growth opportunity. I do believe from a branding perspective, it's an excellent opportunity for us to show off what we can do with technology that may be too expensive to implement in the lower end, so to speak. We will give our customers the opportunity, of course, do you want to sell it? If you want to sell it, we do it together. If you do not want to sell it, then we'll do it one way or the other on our own, but it will not compete with their core business. I'm glad you asked that question.
Sorry. Thank you.
Could you be kind and turn around? Okay. Same questions. That's fine. Perfect. Thank you. Anyone else? Any takers? No. Loading. We'll take those, if there are any. Proceed, please.
I think it's still you, John.
Yeah. We're going to break for 10 minutes according to your agenda.
No, we will.
Just keep going?
Yeah.
Well, at this point, we're gonna break with tradition because we don't have a traditional dog and pony show for you. We're gonna take a slightly different approach. I'm gonna kick off and just provide a status update, and then André is gonna provide some provoking testimony, I'll suggest. Can I say that? Without much further ado, let me get started. As I say, I have a short preamble, and then André gets to do the fun stuff. Where are we with data center? This started back in 2013. I know, I was there. It was a very clever idea, very exciting idea. It was that, hey, we figured out that water's better at moving heat than air, and we've been able to prove it in a lot of desktop computers. Why wouldn't it work in data centers?
Of course, it does work in data centers. Off we went with our new ambition, our new ideas, our new plans. Our report card today, not too bad at all. We have liquid cooling installations around the globe, and that's true of Europe, North America, and Asia Pacific, Japan. In actual fact, as of today, we have 13 of the world's most powerful and equally important, most efficient, supercomputers as listed by the November edition of the TOP500 list and the Green500 list. If you study the TOP500 list, we have three of the top 20. I'm gonna tell you, our stuff does what it says on the box. No doubt about it. We're not happy. We're not satisfied. My guess is you guys are not satisfied either, because we've never really achieved the velocity we had hoped for.
Looking forward, I don't think anyone in the room would argue with me if I suggested that we're gonna continue to see exponential growth in data use, autonomous driving, enough said. I don't think we're surprised that data centers are consuming more and more of the world's energy. Our view is that energy efficiency and carbon footprint are gonna become very important in the data center world. Very important. Our conclusion is that if there were to be legislation, it could have a dramatic impact on our business, perhaps act as the catalyst to get us to the kind of trajectories and run rates we'd always dreamed of. A quick word on one of our most recent innovations, the InRackCDU. We have sampled some 14 customers, 21 units in total. As it stands today, testing is either underway or planned to be underway.
It is way too soon to start making any commercial predictions on how the InRackCDU is gonna impact our revenue. With respect to revenue, you should be familiar with this chart, as there should be no surprises there. We have been able to build a business and sustain reasonable levels of revenue, but nowhere near our ambition. We do continue to execute our strategy. You'll note that in Q1 of 2019, we've actually been able to add a couple of customers. NEC Europe, not to be confused with NEC Japan, which is listed in Q4 2017 as a new customer, and XENON in Australia. Both have recently placed orders with us. You won't have seen stock exchange releases because those orders didn't surpass our $100,000 goal for justifying a stock exchange release. We do continue, guys, to execute on our strategy.
That concluded the status update that I had prepared. Now I'll pass it on to André, who will take it from here.
I had another night moment. I've had several of those over the last five years when it comes to the data center, we probably all have. The fact is, although we have very solid and really great design wins to talk about, it's nowhere near where we had hoped to be at this point in time. That's just how it is. The more interesting thing is to understand why. We have very strong value propositions, and as John just said in Scottish, so I'll translate it to English, that our product does what it says on the box. We have some of the most difficult customers you can imagine in Japan, and they are very, very hard to please. Not only do we have customers in Japan, we have repeat customers in Japan. Why is it that the rest of the world doesn't get it?
I think the most important slide or bullet point, for example, or perhaps on this slide, is that energy savings is not really a big deal yet for the guys we are trying to sell to. As a consequence, let's just talk about the company and the market first. We have decided to discontinue data center guidance for now. We will still inform you when we get a new order. We will still inform you when we get a new customer or if there is something exciting to talk about. I, for one, do not see the purpose of standing here quarter after quarter telling you about 4%-5% of our revenue, why that's not going as planned.
I think this CMU, the whole rebranding effort, et c, I think it's a good time to put a stake in the ground and say, "This is where we are." The segment investments, we are going to reduce those. That is not necessarily related to the revenue situation. That's something we've been talking about for a while. I think Peter will put a number on, but I can disclose already now that we expect to be able to scale down the, let's say, investment to $4 million a year. Is that correct John?
Yeah.
There is an apparent need for standards in the data center market of what you can and cannot do when you build a data center. Therefore, we are going to put effort into try to influence the influencers and try to influence the EU and the politicians.
For those of you who live in Denmark and for those of you who live in Europe, probably also, you cannot open a newspaper, you cannot open a TV show, you cannot open anything without it talking about now you cannot eat meat every Wednesday, or we have to save the world. There's all this about carbon emission. The crazy thing is that one of the biggest problems on the planet is the prediction of data center power usage. Funny enough, nobody's talking about it, because it's difficult and because it's new. If you look at my slide here, the global electricity demand is roughly 20,000 TWh . That's a big number. 3,000 TWh is what they expect to go into data centers to power the data center.
If you look at the recycle angle, 70% of all this power could be recaptured, meaning that you could save the same 70% on CO2 footprint. That means 70% of 3,000 TWh could be reduced. We don't even have cows enough in Europe to match that number. It's really big numbers we are talking about. On top of that, you have the power consumption, where liquid cooling by itself will give you 20% power saving. As you get power saving, you also get CO2 savings. Just to take this down to Earth, we made this slide, which says a lot. In Europe alone, we expect that by a few years, the energy consumption of data centers will be 104 TWh . As I just said, we can recapture 70% of that.
That would be enough to heat 6 million European homes for free and with zero carbon footprint. Of course, there are logistical challenges like where are the homes placed compared to the data center, compared to the district heating plant and so forth, and how many in Europe does actually have district heating. There's more than 6 million people in the EU, so from that perspective, it's a non-issue. If we bring it further down to Earth, now that we have to drive electrical cars soon, all of us. If we look at something that will be much more beneficiary. In Denmark alone, Energistyrelsen predicts that by 2030, it says 2029, but 2030, we will spend 7 TWh of power in Denmark just on data centers. 70% of that can be reused as hot water without the need for heat pumps.
That's as much as the three largest cities outside Copenhagen. It's 300,000 households. It's Aarhus, Aalborg, and Odense. There is nothing wrong or nothing hindering this to be reality. This is not a dream. If you go back five years, in our media room, you can see videos from the university in Tromsø, in Norway, where we are doing this. There's no black magic into it. Why is it then that people are not buying it? My conclusion, our conclusion is that who are the winners with what I just said? It's not HP or Dell. For them, the world is more difficult with liquid cooling because it's new compared to what they do. It's not necessarily the power plant either, because now we have all this free heat, and they are making a living out of selling it.
It's not necessarily Apple or Google or Facebook, because they just want the data center to run, and provided it's run by power from the wind turbine industry, they are green. If they offer, for example, the municipality of Viborg to say, "You can get all our waste heat for free," then they're even more green. The challenge is that the waste heat a data center like in Viborg produces will be 10- 15 degrees cold water. If you put 10- 15 degrees cold water into our radiators, we will cool this room instead of heating it. They came up with this great suggestion that let's use heat pumps. The problem with using heat pumps is it consumes power. There, the green profile kind of went away a little bit.
Number two, it costs millions to install all of these heat pumps, therefore the math doesn't stack up. What we will try to do is to influence the influencers in a way where we get to a standard like we have on cars, for example. You cannot drive your own diesel truck in the middle of Copenhagen because it's not Euro 6. If you have a Euro 6 norm truck, you can drive it pretty much everywhere. I believe, we believe on data centers, for this to take off, we need somebody of the politicians to wake up and say, "Wait a minute. If you want to install a data center in Denmark or in Europe, it has to be efficient. You have to be able to reuse 70% of the heat, and you have to be able to provide it as 60 or 70-degree hot water.
If you cannot do that, you cannot place a data center here." As you can already hear, unfortunately, this is not something we can fix over the next three quarters. I honestly believe this is the missing link. This is why we don't see the traction we had hoped for, because the value proposition we offer is strong. It's on the highest political agenda globally. Here down on Earth, nobody cares about it right now. That's the status as I see it. Before I move on, we can take questions or comments now.
Any questions?
Yes, sir.
There you go.
If that's the biggest impediment, what do you think you can do to accelerate the process from a regulatory or political standpoint? What sort of resources can you put into that to get this adoption cycle a bit quicker?
That's a great question, Scott, because I have been a little bit afraid myself. We are a small company. What can we even do to have any impact? Luckily, unfortunately, I've been meeting with some people who gave me a little bit more hope that even if you're a small company, you can actually do stuff. For once, there is a professor at the local university here, Hemmer, I guess, his name. It's completely coincidence that he's located here in Aalborg, but he is actually one of Europe's leaders within District Heating, and he's been a big part of defining what they call District Heating 4.0 throughout Europe. He's starting up a new research program now together with Danfoss, together with Grundfos, together with LOGSTOR, I believe, some of the really big companies in Denmark who also have an interest in this.
For the first time, he's pulling together forces both from the data center industry and from the District Heating industry. Of course, we want to be part of that. Of course, I want to be a part of that summit. We may even want to sponsor both money and hardware into this research program because I believe that's the way to do it. I also believe that living in our own little bell here, that of course, everybody knows about Asetek, and of course, everybody knows about data center liquid cooling. I had this professor and some of his people here, and they were super excited seeing what we could do. They had no clue. I showed them the video from university in Tromsø that we were reusing the waste heat. That's a part of this research program that they wanted to investigate.
How can we do this, or can it be done without this need for heat pumps? I do believe we can, if we do not use the ugly word lobbyism, then I do believe we can place people in Brussels that can at least figure out who do we need to talk to, because I actually believe a big part of this is really just to tell people about it. I can wake up in the morning, blame the politicians, say they are stupid. Now, Tuesday, I cannot eat beef, but at the same time, I can save the planet. What is it they don't get? What they don't get is that they've never been presented for it, and that's obviously something we can change.
On the marketing side, I have engaged an public affairs advisor who can help us get to the right bloggers, help us get to the right media, help us get to the right politicians. It's election year this year in Denmark, they are craving to tell a story about how you can save the planet. They don't know about Asetek. That's what I believe we can do. Unfortunately, what I cannot do is I cannot say how long time is this going to take? What's the revenue? When's the revenue? That's why we have stopped guiding on it. Not to hide anything, because it'll be completely transparent what the revenue has been. To be honest, I'm a little bit tired of showing up here, telling a great story about our overall business.
$3 million of the business is dragging down everything. Instead of focusing on the cheese, we focus on the hole in the cheese. Again, not trying to hide anything. I believe in this more than ever. I'm just sorry to say that we are still too early. That's the only conclusion I can come to.
Yes, Poul?
Okay. Poul Ernst Jessen , Danske Bank. I have a couple of questions. The first relating also to guidance. If you stop guiding on data center, does that mean that the guidance of 0%-10% actually relates to gaming and enthusiast segment mostly, or?
It's full launch.
Okay.
What we're not guiding on is, okay, how much of that 10% growth come from data centers and how much come from desktop. Again, it's not to hide anything. It's just if I say a number today on the data center, one thing is for sure, because six years of doing this have told me that I'll be wrong no matter which direction it is. Instead, looking at the bigger picture, is the company performing and are we getting traction on the data center side? I think that's the important key parameters.
Yeah. Okay. Looking at the In-Rack CDU product, you mentioned a bit about it, is it still on track to come out this quarter?
Well, you could say that it is out in the sense that I don't think I'm exaggerating. I think the product is ready. I think you can actually, when you go out, see them building it down here. Of course, it's up to the customers to fit into their schedule when they want it and when they are ready, et c. I think I promised in Q3 or Q4 that we would be ready to launch in Q1, and we are for sure.
Have you received any orders already, or?
It's very modest. Evaluation quantities, two or three. It's nothing that would constitute what we would be looking for in terms of production volumes.
All right, great. Thank you. My last question, you talked about scaling down the investments in the data center segment to $4 million a year. Can you just tell us a bit more how you're going to do that and what it entails?
That's actually fairly easy because we'll just stop developing new products until we have customers who want to buy. In other words, I believe we have the technology that's needed. This is not a question about let's build something that's even more efficient, then we will get the orders. We have that. What we need to invest in still, of course, is that whenever Intel or NVIDIA comes out with a new platform, for example, Intel is coming out with a new platform later this year. Of course, we need to make sure we can support it so we can actually take orders. Of course, we need to keep the lights on so that when customers call us, that there's actually people in the seat to pick up the phone, and et c. So that's where the $4 million number comes from.
What we are not doing anymore, even who it is, even if it's Intel calling again, there needs to be a solid business case, or we will focus on our core business and focus on that let's get some rules around data centers. Just back to the data center in Viborg, of course, Apple would not have said, "Oh, if that's your demand, we are not going to place a data center in Denmark." Of course not. It's not more expensive to build it with our solution, but it's just as long as nobody's demanding it, why would they do it? Huh, I just lost that. Anyway, what I wanted to say is, of course, we cannot get EU to say, "You have to buy liquid cooling from Asetek, or you cannot place a data center in Europe." That's not how it works.
There are other people who can do it, not as good or as efficient as we can, I believe. Of course, there are other people who can do it. I do believe this vision is real, and I do believe we should do it. I do believe that if they are serious about reducing the carbon footprint, that they should look at, to use your words, the elephant in the room, because right now, just as it is, a big part of the green power in Denmark will actually go to data center. Then you can say, "Yes, the data center is green," but what about the power the rest of us get? I believe this is real.
Thank you. Sir, Mr. Knudsen.
Talking about elephants in the room, the In-Rack CDU, when do we know when it's a success or a failure? How long time are we going to give it?
I'm probably going to give it less time than you. Let's put it like that. It will not be me you will be hearing dragging this out. We want to find out, and we want to find out as soon as we can. Of course, we need to give our customers a fair shot at actually testing it. Although I would hope, then In-Rack CDU is not necessarily as high up their priority list as it is on ours. This is not something that's going to be dragged out. You can count on that.
Very good. Anyone else?
It was my impression that on the data center side, it was twofold. The energy efficiency and then also the CPUs are getting hotter and hotter. I think that's still the case, right? When Intel is bringing new CPUs to the market that will demand liquid cooling at some point, or has that case also changed?
It's not off the table. I think we have the winds in our direction, so to speak, on all these fronts. However, we believe that for six years, and what we learned is we are not very good at guessing when. I do not believe it's a coincidence that we have Japan as perhaps what our biggest market.
Yep.
Why is that? They had a little accident out there a few years ago, power is now a limited resource. Power prices go up like crazy. Therefore, actually being able to reduce power is something that means something to them. In other words, as soon as we see customers that like our value proposition, they will buy. I believe that's why we are successful in Japan. It's not because of price, that's for sure. Our solutions out there are fairly expensive.
Very good. Anyone else? Here. I have a question from Germany. How do you see your competitive position in data center has developed recently, also on the technology side?
Lots of great questions today. This one I couldn't have made up better. That's a really great question. If we had competitors out there that would be as successful or even more successful as us, I would be happy, we are not the only idiots on the block. The fact is, I do not see any competitors out there who is actually making a great job in actually getting business. I think most of us know that Google has made a data center where they have deployed liquid cooling. The challenge for us is that Google also came to us, but what they wanted to buy is a piece of copper with two pipes sticking out of it, and if we are lucky, we also get to sell two pieces of rubber hose. We cannot make money on that. Asetek is not a metal bender.
We are not a cold plate manufacturer. We are a system supplier. We are not that desperate for revenue that we will ruin our core business and end out being a 10% margin company just because we are so eager to get progress on the data center side. I guess the short answer is that we are in this together with our competitors, I do not believe being a component vendor aside, that we will succeed until we actually gain traction on the politics side. That could, to your question there, Scott, be another solution that we actually team up with some of our colleagues in the business to apply a bigger pressure, so it is not all on Asetek. That's seen before. For example, from the wind turbine industry.
I believe both Siemens and Vestas have been together in, again, let's not use the lobby word, but in pressing the right dots in EU to get the focus.
Very good. Any more questions from here as of now? Seems not to be the case. Do you have any more slides there? I assume not.
No.
Let's take a 10-minute bio break. There's coffee and whatever you have over here. Let's meet back here in 10 minutes. Thanks.
[Break]
Last cups of coffee. Take your time, John. Very good. Very good indeed. We released our Q4 2018 report and our annual report 2018 this morning, both of which we are very happy about. Bottom line, pre-tax for the quarter alone was $1.2 million in 2018 versus the same quarter last year of only $500,000. What's not to like? On the top line, over the last five years, we've been growing 26% on average. That's nice. That's good to look at. Let's make sure that we continue that by putting more resources to the Gaming Enthusiast so we can keep on growing that business. For the year 2018 as such, $67 million, which is a 16% increase over the year 2017.
Most of that, of course, comes from what we now call Gaming Enthusiast, GE, formerly called Desktop, with $63 million, which is 18% increase versus the year before. We made $9.4 million worth of EBITDA in 2018. That's a 38% increase. 2018, 16% up in revenue, 38% up in earnings. That's an okay picture seen from my chair. On the right side of the graphics here, what we're looking at is, of course, the group revenue, as John alluded to Q2, that one here in 2018 was an extremely high number, Of course, that makes our comparisons going into 2019 a little bit strange, We'll have to take that one quarter at a time as we go along. How does my world look? We need to build continued profitable growth.
We've been talking about that. That's our aim. Of course, we need to do that on a solid financial platform. If we break that down into what we call value drivers, on the gaming enthusiast side, desktop, I have to get used to that also, GE, gaming enthusiast, how should we find the words? We'll figure it out as we go. We are rebranding and strengthening to strengthen the market position. We need to get closer to our end users, getting closer connection with our end users, thereby, of course, grow the revenue. This is not an exercise of changing a logo, even though the logo looks nice and the Facebook is going to look nice, I'm sure this is not what the focus is all about. The focus is to maintain our growth position on that side of the business.
We're also working with the diversification of our revenue streams. John, I believe, talked about it, how the split between our top five customers look more and more healthy. We used to have one very large customer. Now we have five, even six larger customers together. Of course, that's an exercise that we need to continue to work on. It says margin protection and optimization. Fancy words for making sure that our margins are at the level where they should be. Gross margins in the last many years have been around 40. We've actually seen gross margins recently up above 40, but 40 is a good value stick to keep track of. On data center, the next box down, we're talking about maintaining our data center market position. That's about ensuring our operations are efficient and making sure that the adoption continues to happen between our customers.
Yes, we're not going to focus so much on it as we have done. We're still going to sell the products. We have a nice palette of products. Of course, we're going to sell it. That's what you're going to see in the future. It's also about optimizing the cost of goods. We have a fairly new machine, infrastructure manufacturing machine that we still need to fine-tune. We've seen a lot of new products over the many years. You've seen margins, gross margins go up and down because cost of goods have been going up and down. That's a continued exercise that we do to optimize our cost of goods. It's about cost-based optimization. In other words, overheads. That's like an inwards kind of exercise.
We need to make sure that we spend our money in a way that we can talk to you with our eyes open and be okay with the way we spend our money. You will see when we get a little further down that our operating expenses from 2017 has been pretty much stable into 2018. That's a deliberate choice we've done. We have followed our policy on that. We could, of course, develop all kinds of new products and sink a lot of money into R&D, but we have a good debate ongoing in the company about how we spend our R&D money, focusing on the right projects. That's an ongoing discussion. Manufacturing, we also have to make sure that we build the machine, as I call it, here locally and/or in China, or if we buy or make the products ourselves. That's also an ongoing discussion.
It's about sales and marketing efficiency. Last time we met a year ago, we pretty much didn't have a marketing department. That has changed now, but that's also a deliberate choice. We need to beef up our muscle to be able to go and develop new products, yes, but also sell it and market it. That's a deliberate choice. Cash flow improvements, ongoing exercise, converting sales and purchase into cash in the bank account. I'm not totally happy with the way we have performed this year on that particular topic, but I'll come back to that. Revenues. You've seen those slides here a couple of times today. We see a growing business. You also see a fluctuating business, as André, he said, we always have one bad quarter. We just never know where it is.
John mentioned into saying that it's often Q2, but it does fluctuate up and down. That's just how we live. If we, in order to flatten that out a little bit, look at trailing revenue instead, then you will see a different kind of picture. You'll see a growth case. Again, 20% increase in revenue constantly over the last five years. We started making money, bottom line, in mid 2015, and we're actually making pretty good money, you could say. That's the interesting part here. If you look at the EBITDA in dollars, going all the way over here, $9.4 million. That's just the nature of the business here. If we have a growing business, we have a fairly stable gross margin, and we have a fairly stable overhead cost picture, then earnings will increase when revenue increases. Let's back to growing the business. Gross margins.
This is a little bit difficult to see maybe, but on the left-hand side of the screen, we have the last five years of gross margins. Let me get data center out of the picture first. That's the gray line that you can see going up and down. It has been fluctuating quite a lot. It has been improving a little bit in 2018, but nothing significantly. Again, it's a small part of the business. It's so small that we do a lot of one-offs, and it's so small that by nature, the gross margin will fluctuate up and down as is the case. The blue line, the black line is the rest of the business. You can see over the course of the years, we have been around 40.
There was a little bit of a hiccup in 2015 where we had some one-off expenses, but apart from that, it's 40 and it's increasing. We had some headwind in 2017, pretty much driven by currency exchange rates that we have now seen coming back to help us in 2018 instead. We have actually, if you look at the slides on the right-hand side, where it's the last five quarters, then you can see that we've increased our gross margins on the gaming enthusiast side of things to above 40, helped by a sales price increase that you may remember that we implemented in Q1-ish of 2018, but it takes some time before it actually rolls through.
At the same time, when we implemented the sales price increases, we also turned around to our engineers, asked them to develop products that were more cost efficient, we turned even further around, talked to our suppliers, and asked them to help making sure that we got the best prices possible. All in all, those things have come together during the Q3 and Q4 of 2018 and has given us fairly nice gross margins. Going forward into 2019, if I was the guy with the Excel sheet, I would model 40 as the marker for gross margins on the gaming enthusiast side of things. Especially now when, I will come back to that a little bit later, with the currency exchange rates on China have increased a little bit in recent months.
You never know what's going to happen there. One more component of the increased improved gross margin is the fact that we see sort of a slide towards more complex products, higher end products that drives up ASP, yes, but also drives up earnings per unit. Here was my currency exchange rate slide. I do not want to spend a lot of time on it, the one on the left is USD versus renminbi. We saw a big increase in the price of the Chinese currency in 2017. That's what we have over here, 6% up. All of that came back and helped us throughout 2018, and that was nice of course. We now see an increase of the renminbi, bringing us up to the level around July 2018.
There is some time lag in when these things impact our gross margins, there is an increase, that is why I would be nervous about projecting a gross margin much higher than 40 in our case. Add to that at least I have no clue what is going to happen on the political scene when it comes to the currency exchange rates between U.S. and China. It's a battlefield. On the right side of the slide, I am showing the exchange rates between DKK and USD, I do that because 80% of our overhead expenses are denominated pretty much in DKK. Pretty much 80% because we have the facilities here in Denmark. We saw an increase of the cost of running this operation over here in 2017, half of that has come back and helped us in 2018.
Net currency cross seems a little bit more stable at this point in time, again, you never know. If we go further down, we have talked about revenue, we have talked about gross margins. Let us talk about EBITDA. Let us start with the gaming enthusiast side of things. It is a profitable business. It is an even increasing profitable business. We have an EBITDA percentage here in Q4 of a whopping 35%. A very good business. We have not grown as much lately as we would like to do. Hence, let us go back, invest in the segment. Let us make sure that we start growing that again. We can afford that because see how, again, the EBITDA margins are going up to a level that are quite high.
If we take, for example, the $4 million that we're saving on the data center side of things and invest that amount or less in the gaming enthusiast side of things to enable the increased growth, then we should see continued earning also here. Data center earnings, same picture as you've seen many years by now. We have been focusing, we've been putting quite a lot of resources into this segment, some quarters more than others, simply depending on how much engineering resource we are pouring into this area here. Right now, this exercise of moving resources from data center over to gaming enthusiast, it's an ongoing business, ongoing enterprise.
What happens is that we have asked our research and development department to simply look through all the lists of actual projects, close down what is not necessary in the new picture here, and then start focusing on the enthusiast and gaming projects instead. It used to be, again, for us with the Excel sheet, it used to be that we spent 60%-70% of all our resources on the data center side. The march order here now is to go down to reduce that to between 20%-30%, and that's going to happen over the summer of this year. It is very much an ongoing exercise. Combining the two numbers, I just want to remind you that, yes, we have been spending money on the data center business, but we have also been earning a lot of money on the gaming and, what do we call it?
Enthusiast. Gaming and enthusiast. I will have to remember that and exercise during the night. Going away a little bit from the trends. I know there's a lot of data here. On the left-hand side, we have 2018 as a full, we have 2017 as a full. We're breaking it down between the gaming and enthusiast, I made it, and data center. We have, of course, starting out with the revenue top line, which we talked about growing that from $58 million- $67 million. We talked about the gross margins that have gone up from 36%-39%, all a nice business. You can see here the total operating expenses, pretty stable from $12.2 million in 2017 to $12.7 million in 2018. That's a 4% increase. Keep in mind that we increased our revenue by, was it 16%?
It's a fairly flat number, I would imagine, or I would claim. What's important here to look at, you can see smack in the middle. Here we go. You may not be able to see it on the screen. Total operating expenses, gaming enthusiast in 2018 was $4 million. That's what we want to expand. We want to make sure that we can grow the business, and we do that by decreasing the data center spend to $4 million, as André, he was alluding to before. That's the end goal is going to be reached at some point during 2019. The split of resources that we talked about before, going from 60%-70% on data center to 25% data center. The yardstick has been set at, I think it's July 1st or August 1st, or something like that. That's the goal point.
Of course, there will be a transition. This is a year of transition, the idea is then to, again, spend more money because we need to spend more money on developing products, markets, et c, on the gaming enthusiast side in order to, not in 2019, don't expect that, but in 2020 to increase the revenue on that side of the business. Again, it may look like an Excel sheet that we're pretty much swapping these two numbers here from $8 million-$9 million of data center spend to over to $8 million-$9 million of gaming enthusiast side. This is a very real exercise. It's real people we are hiring. We have hired a marketing department. We're also letting people go, and we are reassigning people from one project to another. It's very real.
It has happened over the last months, I'm sure it's going to happen also in the future. This is by no means an Excel exercise. Enough about that exercise. EBITDA, $13 million this year versus $8.7 million in 2017. Depreciations have gone up. That's again, the nature of the beast when you're in an R&D business. When you are done with projects that you have capitalized, which is what we have to do, you have to write them off. We have had some quite significant projects in recent years that we are then starting to write off here in 2018. Keeping in mind that our write-off depreciation period is relatively short on the gaming enthusiast side of things. It's 18 months, so it should be off the books fairly soon. What else to talk about? EBIT margin 13% this year versus 8.8% the year before.
Litigation expenses, that's always also an interesting topic when talking to you. Has been fairly flat between 2017 and 2018. Last year in 2017. Excuse me, when I talk, I'm an accountant. When I talk last year, it's probably the year before. It's 2017 in this case. Last year, 2017, we received a settlement income of $900,000, which of course we didn't do also in the same in 2018. When you look at the bottom line numbers, which were in 2017, $2.7 million of EBIT, then you could actually argue that we should add that $900,000 or we should deduct those $900,000. Instead of $2.7 million, it should say to $1.9 million of earnings in 2017, and we are now at $4.4 million in 2018. That's an increase of 140%. I think it's quite respectable.
There's a reason why we were happy this morning being able to send out the annual report and the quarterly report. I should add, because otherwise you will believe that I'm cheating, there's one line here called HQ other, and in hindsight, I would have liked to show that in a different way this year. It's mostly a matter of us being more diligent when we look through the expenses over the course of the year, whether it has been a gaming enthusiast expense or data center expense, or an HQ expense. Live and learn. We've been better at doing that. As such, there's not a big increase in the HQ cost. However, we have spent some money, of course, on consulting and strategizing around this new set up there that we are presenting today. Bottom line, the EBIT $4.4 million. Yeah. Cash.
Cash is always an interesting topic also. The way this slide looks, let me just explain. On the left-hand side here, we have a black bar saying $ 13 million. That was the cash we had in the bank at the end of 2015. We then made $ 15 million in EBITDA on the desktop, spent $ 5 million on data center, then other expenses ending us up at $ 17 million at the end of 2016. Same in 2017, the same in 2018. What has changed, apart from, of course, the numbers being bigger in 2018 is not a lot. That's coming back to, if there's one thing here that I'm not proud of, then it's our working capital managing. We had the same in the bank at the end of 2018, $ 18.6 million, pretty much as we had in the bank at the end of 2017.
With a profit that he just bragged about of $ 4 million, how can that be? That is primarily an issue, a matter of us giving a little bit longer- terms to our customers and us not getting the same long- terms with the vendors. We have it right here. There are three key performance indicators that are important when it comes to cash in our business. It's inventory terms. Not that our inventories are high because most of our sales go directly from our factory to the customers at the point of our factory door. Still inventory is going up from 2.3- 2.8, that means that we were a little bit less good at managing our inventories in 2018 than we were in 2017. They turned around 23x . In 2018, they only turned around 14x .
More importantly, on our receivables, up from $ 13 million- $ 15.5 million. Doesn't sound a lot in the big scheme of things, in any event, it's a change from us giving terms of 64 days on average to now us giving terms of 79 days on average. That's a 15 days longer- term on average. That's quite a lot. Combining that with the trade payables, where we in 2017 on average had terms of 77, we are down to 67 in 2018. We're not as good as we used to be at getting terms from our vendors. I'm speculating here, this could actually be where we see the effect of the U.S. tariffs, because what happens here is on the receivable side, it's of course, it's a debate with our customers.
Yes, Mr. Customer, we would like to charge you a higher price, the customer comes back and says, "Well, okay, you can charge a higher price, but we need a little longer- terms to pay them." That's a very common discussion. The same on the trade payable side. We are turning towards our vendors and say, "We need you to pitch in and pay a part of the tariff cost here." They will say, "Yes, but then you have to pay us faster." This may actually be the picture here. I think things are improving a little bit, it's certainly our intention to have the cash conversion come down from 39 days towards maybe a zero. That's our big fat round target goal. Balance sheet. You have seen this slide a million times. We are a cash-rich, very solid company, it's on purpose.
We need to be cash rich and very solid because when we talk to new customers, they want to make sure that we're here tomorrow too. When we talk to new vendors, they want to make sure that we are here tomorrow too. It goes round and round. Also, if we talk to competitors, we want to show enough muscle that should they choose to go into a lawsuit with us, we are then solid enough to withstand such a lawsuit. There has been some debate, well, debate, that's maybe discussions over the last year or so about whether we should pay out a dividend of some sort. Those of you who's been with us will know that we went through a dividend payout in the year of 2017. It will haunt me forever.
I grew 10 years older in two months or something like that because it was quite complex. It was for a reason. The reason is that we, as the parent company, has to be considered a tax resident or tax citizen both in U.S. and in Denmark. It's certainly not a desirable situation. It's not a situation that we wanted to be in. It's the case that when we moved the company from U.S. to Denmark some years ago, back in 2013, 2014-ish, the U.S. was very focused on making sure that people didn't act like Apple, for example, who moved operations to a low-income tax country.
They pretty much put down the shutters and say, "If you move out of U.S., we will still consider you a U.S. company." We had thought we could just get through a little hole saying that if you move back to your own country, that used to be the rule. If you move back to your original company, then we wouldn't have that situation. They closed that hole in 2013 on us. All this is not a problem for the company per se, but it is a problem for the investors, particularly investors who are not U.S.-based, and particularly for the European investors who are small and don't have a presence in the U.S. If you have a presence in the U.S. and can show me a form, which is called a W-8, I believe it is, then you're home free.
Not a lot of people can do that, and that puts us in a fairly complicated situation. That means that coming back to a dividend, meaning that dividend payments is, in the current situation, not something we can do. We turn our focus to a share buyback program. Could we do a share buyback program? It then turned, and people said, "Oh, they do that in the U.S. all the time. That's not a problem." If you start looking into it, and I can assure you, we had some very competent people from different firms looking into this very simple topic. It then turns out that the Americans consider a share buyback program to be similar to a dividend payment, unless a certain number of things are checked off the box, off the list.
At this point, we're not able to check those things off the list, meaning that even if we do a share buyback program today, for those who are foreigners, as seen with U.S. eyes, meaning Danes, Norwegians, what have we, they will have an issue with a share buyback program as we talk today. It's not a very appealing situation. It can be done, and we are going through the exercises, the motions to make sure that it can be done. For now, it is also considering the fact that our cash balance has been fairly flat for a while. We're looking into at least half a year where we are very uncertain about we see a decline in revenue, at least for Q1, right? For now, the plans of dividends and a share buyback program has been put to a side.
We're setting ourselves up in a way so we can pull it out of the hat and actually exercise that fairly quickly. We are discussing whether to work with the Danish tax authorities and the U.S. tax authorities to solve that conflict that they have. It's written in a double taxation treaty that there should be no double taxation going on, and this is the case, there is double taxation. Please, authorities, please fight it out, and we'll of course Not of course. We're considering whether to have them do that. The outcome is not clear at this point, though. Good. Oops. That was the pointer. Ending up here with the financial outlook, the same way as I started out with the priorities, continued profitable growth, and a solid platform.
We have seen growth, we will facilitate even more growth, and we have seen a 2018 with a very strong bottom line work in adding to our solid financial platform. The question is here, how to continue, what to do now? On the revenue side, we are guiding our revenue for the total group of between 0% and 10% compared to last year. You know, as I know, that there are some macroeconomic uncertainties that are tempering the whole expectation set here around gaming enthusiasts. We also talked about how we are protracting our expectations on the data center side of things. Margins continue at current levels. Be positively surprised if we are above where we are today. On the data center side of things, they will keep on fluctuating as we have seen them. The trick, the key here is scale.
When we get to more scale, our gross margins will be more stable and will increase also. Caps and allocation. R&D, we've been spending between $3 million and $5 million in the last couple of years. We will do that going forward. We will move R&D spend from data center to gaming enthusiast. If there are cost savings to be found, we will take them. For now, for modeling purposes, understand, expect that we are moving from one segment to another, and this is a year of transition. CapEx, $2 million-$3 million, expect it to be in the lower end of that range because I don't see we have a huge need for further enhancing our infrastructure. Headcount, I think we are 100 people right now. Expect us to be at that level also for 2019.
In all, expect a modest growth on the overheads at group level. If there are savings opportunities, we will take them, but don't build it into the expectations for now. Financial position flexibility, yes, I just talked about it. We will continue being as a cash-rich company, and we will continue, of course, to come back to working on our working capital to get hopefully the tariff issue fixed. All this in total should, cross your fingers, knock on wood, appreciate the share price. Good. With that, should we see any questions for now? Otherwise, we have a summary and outlook. We actually do have a couple of questions. Sit down, Mr. CEO. One from Germany, how much your reduced investment in data center could burden your cooperation with Intel?
You don't want to answer that?
Like, it-
I don't think it has an impact.
Right.
Number one, we already developed the products, number two, everything I said today that's on the next slide does not mean it's true. If there's a new opportunity coming tomorrow, of course we will look at it. We are not trying to create a self-fulfilling prophecy here. If a new data center opportunity opens up and we consider it valid, we will probably do even more homework than last time, of course, we will look at the situation and see whether we need to invest again or not. For sure, by reducing the expenses on the data center side does not mean that we have closed any doors. A little bit provocative, I can say what it means is that we have shot the free shot of R&D.
I think that's probably the best way.
Yep. Same gentleman in Germany is asking, any hopes, and I like the word hope here. Any hopes for lower legal costs going forward? Of course, there's a hope for lower legal costs. It's not that we are calling the attorney asking for new things to do. What do you say, André?
Well.
You have a couple of cases going, right?
I have two answers. One answer is that the current case that is progressed the most against AVC, they actually sued us to get a verdict that they were not infringing our patents. It is kind of the other way around. Of course, there is nothing we can do to not incur that expense. My hope is, again, good and intelligent question, my hope is that if some of you remember back to my presentation on the IP, I had the branding. I actually do hope, and it is a part of the plan that perhaps we can spend the same money or more money on the branding and save them on the legal expense. I believe that Branding, at the end of the day, should be stronger than your patents.
In reality, it should not be a patent that decides what product you buy in the end, it should be the brand preference that you have. Yes, I hope, but I cannot predict what our competitors are doing.
No.
What I would like to say also, though, is we will continue to sue people who violate our patents.
Yes.
That's our obligation. That's why we have the patents.
Yes. Time has shown that defending your patents and paying for it is a part of the price of having a high gross margin product portfolio.
Yep.
Any questions from here at this point? Otherwise, summary and outlook.
Yeah. I'll do it short because we've recently been through it all. We are the market lead, and we are de facto being realized and recognized as the industry standard. We are now evolving our brand to actually support that statement, and we'll do that by looking more into the gaming and the enthusiast market and increase the awareness about Asetek among the end users. I believe we are still in the forefront, very forefront, still early stages of the data center business. Unfortunately, however, the money we have spent are not wasted because when the market kicks in, we have developed it. Yes, from a timing perspective, the situation could have been more ideal, but it is what it is. Following the strong 2018, 2017, 2016, 2015, we do look at a more tempered outlook for 2019.
At the same time, I would also like to say that looking at history, I try to be conservative, and I do not know what's going to happen. We could get a positive surprise. We could also get a negative surprise. I don't know. What I'm relaying to you is exactly what our customers are relaying to us. First half is going to be weak for many different reasons. Second half, things seems to pick up again. That's what I'm telling you, and that's what we have been translating into a ceiling of 10% growth. Of course, we hope that we can beat it, but it's just too early to say. Thank you. I don't know if there's more questions than answers. Let's open up for it.
Any questions from here? Counting one.
Yes, sir.
Just on the data center side, what do you have in backlog right now? I know you're not guiding officially on that segment, but you announced in January a $ 600,000 contract to be done in Q1. There's a few deals from the past that are still going to contribute in revenues. Secondly, at the Capital Markets Update last year, I think, John Hamill, you mentioned there were around 30 potential sales in the pipeline-
Yep.
Varying degrees of probability, high, medium, low.
Yep.
If you could update us on the sales funnel from there, based on your discussions with all the various OEMs. Thanks.
First of all, our backlog is dominated by the Fujitsu order that we announced. I think the question is, how much of that can we ship in this quarter? We received that order in late January, which was right on the cusp of Chinese New Year. That meant that we couldn't communicate with our supply base, so we've been running into some challenges with respect to shipping all that product this quarter. It's a situation that's been exacerbated by the trade war, because rather than take two weeks off as is typical, a lot of companies shut their doors for three weeks and even longer. We kind of walked into a bus sort of there. That's on backlog. It's just whether it ships in Q1 or Q2. In terms of our pipeline, our pipeline's probably the weakest we've seen it since we started building our pipeline.
We've actually been very successful at taking opportunities off the pipeline. You've seen them. You've seen them show up as orders and stock exchange releases. Some of them, we don't make stock exchange releases because they're below our threshold. We've been pretty successful. What we've noticed recently is there hasn't been new opportunities coming on behind the opportunities. As the pipeline matures, there hasn't been as many new opportunities showing up. I don't have a canned answer for why that might be. I can tell you what Fujitsu's telling us. What Fujitsu are saying is that Intel has a new technology coming out later this year, and the market's holding its breath waiting on this new processor. I don't know how to validate that, guys, but that may be one explanation for why our pipeline is as weak as it is today.
Just one more question as well, then, on the gaming and enthusiast side. We know you only have four to eight weeks of visibility, and you've commented that the first half is going to be very weak, yet you're saying you're confident that the second half is going to be fairly strong. You don't have the visibility. What's giving you that confidence, then?
As I said, it's the customers. I participated CES in Las Vegas myself this time and met with the CEOs of pretty much all our customers, and they are saying the same. There is a dimension here that we haven't touched upon, but I also believe is fairly easy to understand. You saw our Q2, how big that was compared to usual. It was almost double. A lot of people bought to inventory to save money because of the perceived taxes or in the trade war. People build up a lot of inventory, and when people build up a lot of inventory that they don't get rid of, then of course, they are not buying from us again.
It's not only gut feeling, it's also the fact that some of our customers said, "Okay, we didn't know where this trade war was going, so we were conservative and bought some inventory, and now we are sitting on some inventory. It's not all just, what are the end users doing? I would actually say, correct me if I'm wrong, John, we've not synchronized this, but I would actually say that, in my opinion, we do not have an end user issue here. We don't have an issue where people are not buying PCs or building PCs or playing eSports. Most of this, also the trade war related thing, is with our direct customers. Their focus is just all over the map, because liquid cooling is not really the big deal for them when talking trade wars, but it's their entire business that's impacted.
Yeah.
I think the softness we are seeing here is something that comes from our direct customers more than the end user. Do you agree, John?
Yeah. It's a function of confidence. Of course, not all customers are the same. Some are impacted more, or behave as though they're impacted more.
Yeah.
Others are less sensitive to what's going on around them. It is a confidence game. We've had customers, I'll share one horror story, just to give you guys a feel for what I mean when I talk to confidence. At the, probably the first half of last year, the cryptocurrency boom was in full swing. We had one customer who couldn't build enough high-end power supplies. These high-end power supplies are needed to power the graphics cards, which unravel the algorithm that allows you to print your own money, believe it or not. Almost overnight, the market crashed, and this customer was left with a mountain of high-end power supplies. I mean, a mountain. Like, they went from being sold out to having two years of inventory overnight.
I get a phone call telling me, "Stop shipping because I can't pay you." Nothing to do with liquid cooling or my customers. Everything to do with a customer having a crisis of confidence because of cash flow issues.
I think another point here is that a lot of people have compared us to NVIDIA in terms of trends.
I think it's important to understand that, yes, NVIDIA is selling into our core markets, for example, gaming and enthusiasts, but they are also selling into other markets, for example, cryptocurrency. I think that's the best answer I can give you when I have said earlier that you cannot just compare one to one. Number one, I believe that that whole deal is inflated. Number two, we do not really sell into it. Yep.
Very good.
Maybe this is a follow-up on the outlook. Have you made progress in Q1? Is March better than February or January? Do you believe this inventory buildup is mostly true?
Yes. The answer to your first question is yes, we see progress in Q1. That's good. I will take it. It does not get us to where I would like to be. When I say the inventory situation, that was a situation that occurred last year. I think we are through the worst of that, absolutely. I genuinely believe the issues we are facing now are more confidence related. Are customers feeling good that this second half pickup that everyone is talking about is going to happen? Are they going to start placing orders as a consequence?
You also, in 2018, talked about component shortages.
Yeah.
Perhaps you can give us an update on where we are there.
Yeah.
That's the opposite of the inventory.
Intel have been plagued with component shortages. That did have an impact on the marketplace. There are varying rumors, depending on where you read and who you want to believe, that they have resolved it, they haven't resolved it. I'm not an industry expert. I'm not watching this day in, day out. As pervasive as it was last year, particularly on the second half, I'm not seeing as many news articles. That's the most specific I can be.
Correct me if I'm wrong, but you had some component shortages, right?
Oh.
Have that been resolved? Not Intel's, but your Asetek's.
Oh, sorry. I thought you were referring to Intel component shortages. Those continue. Small ICs, there's been a chronic demand for small ICs, We're having to go out and forage for units in the marketplace. Those are ICs that we use in the pumps, for example. We have been challenged with that, and as far as I'm aware.
I think it's important to stress that those shortages are related to sudden orders coming in.
Where we have historically, and that's a big part of our growth story, actually, we have had a very dynamic supply chain that's been able to absorb very large orders coming in with very short lead time.
Since we do not carry inventory, or at least a minimum of inventory, then we get to those situations. For our forecast, as far as I'm concerned, I don't think we've had any shortages. It's really this being able to react like this.
I think we're in a situation right now, we could ship more in Q1 that we have plans of shipping if we were able to supply. Let's put it like that.
Go back to your question about how is it looking.
Very good. Anyone else from here right now? Cool. There's one question from Norway. Can you comment on the general cyclicality, that's a tough one, cyclicality within consumer gaming spend? Is it overall linked to GDP growth or less correlated?
I think the short answer is no, I cannot comment on that.
That's a fair answer.
Yeah, I don't have any data.
Nope. Very good. Anyone else here? Going one, going 2x, third time. All right. With that, thank you for your attendance here. It's been a pleasure having you around the world, too. Thank you for your interest in Asetek.
Thank you.