43 min read

EGD 069: Thor Gunnarsson, Mainframe Industries

Today I’m talking with Thor Gunnarsson, the Co-Founder and CEO of Mainframe Industries. Thor and his team started Mainframe in 2019, to build a cloud-based MMO game that you could play on any device. They’ve attracted lots of interest from investors and are now in full development mode. With Thor, we talk about his founder journey, how technology is playing a role in game development and what the future of cloud gaming could look like.

Takeaways

Why did you get into gaming?

“I think the games industry for me is just super interesting because it kind of brings together a heavy focus on technology, and technical innovation, but also creativity. And it’s sort of that combination, creativity, and tech, that for me is just incredibly interesting.”

“And as a guy coming from the business side of the equation, I feel like I just have the best job in the world sitting next to some of the most creative, insane game developers in the world and helping augment make their visions of success”

Can you talk about the game you are making?

“We are creating a sandbox MMO game that has a very heavy emphasis on the social layer and the set putting, creating the setting and the game environment that allows a social network to form within the game, which is sort of a feature of sandbox MMOs that we saw with EVE Online”

Why cloud gaming?

“Cloud gaming engaged streaming is one of the technical innovations that allow us to create a world that people can access, where they can find a window to the world, whether they are playing on their mobile phone, on their laptop, or sitting in front of on the couch in front of your TV.”

“We believe a virtual world that you can access throughout the day on any screen is so powerful because it really, really strongly emphasize the social aspect of your game experience.”

“it’s the same game. It’s the same visual experience. It’s not kind of a watered downside application of the main experience. It’s the same game as all of these fields.”

“This idea of seeing a person playing the game on Twitch or on YouTube or Facebook gaming, and receiving an immediate drop, allowing you to jump into the game, join the game is incredibly compelling from an acquisition perspective and from an engagement perspective”

How do you see a cloud-based game reach product-market fit? What is PMF in cloud gaming?

“First, playtesting: Closed alpha moment, where we’re able to test our core loops, we’re able to bring the first few hundred, first few thousand players to the game very early on, to judge and sort of understand and follow what they’re actually doing in that test.”

“Second, UI/UX: [To have a] PC or console experience that delivers an experience to mobile devices and phones that feels native to the platform that is accessible and can support the kind of shorter gameplay sessions that you would see on mobile typically, and allows players to feel that they can have a satisfying experience again, which is not sort of second class experience to what they see on PC or console.”

LTV calculations for cloud gaming? What are the cost structures of running a cloud game on AWS or something? Or do you need something specific to achieve a positive ROI?

Historically, new tech rolls out slowly and the costs were high early on. 3D acceleration, “broadband capability took a number of years going stretching from first experiences on a 44k dial-up modem to a broad kind of swathe of people that actually had decent broadband access.”

“Google, Microsoft, Tencent, and Nvidia, are the companies that are building out the infrastructure that brings down the operating expenditure. And really, that’s measured in minutes a stream or hours of streaming. And there’s a certain number that you have to reach to make the whole exercise economical.”

“If you’re viewing it as I’m a game developer coming to a platform like Google Stadia like a project xCloud, you are assuming that the cost structure is something that is effectively being amortized, or carried by the platform rather than you covering the permanent cost of that stream.”

“We assume that economies are a scale will kick in and that the infrastructure in the data centers that companies like AWS, Google and Microsoft have around the world, obviates the need for a game developer care as much about the unit economics compared to if he were he or she were running the full technical infrastructure themselves.”

Tools and middleware for clouding gaming are missing. When do we see those come up?

“There are opportunities for middleware companies to create discrete parts of the stack that are not being served adequately by Unity or Epic.”

“In our studio, we have a very specific style of game that we’re creating. It’s a large scale MMO. So we are doing a little bit of internal tech on our site to enable that on top of Unreal Engine.”

How did a16z and others get the conviction, that now is the right time to jump into cloud gaming?

“In the case of Andreessen Horowitz, they are doing increasingly big investments in the games sector, we were aware of them and one of my co-founders Sulka Haro had a personal relationship from the past days with Andrew Chen, partner a16z. And so we were able to set up a very quick initial conversation, it became evident in that first half-hour call that there was a strong match in terms of how we were viewing the industry and the opportunity and how they were had come to their own conclusions about the future of gaming.”

Transcript

Joakim Achren 1:45
Hi Thor, welcome to the show.

Thor Gunnarson 1:47
Hey there great to be with you.

Joakim Achren 1:48
Yeah, likewise! So Iceland. You guys are ending the summer right now, getting back to work. Do you have the same routine as in Finland that July is kind of like closed and then everybody’s like that. First week of august now we’re getting back. How does it go?

Thor Gunnarson 2:03
It’s pretty close. We’re not quite historically quite as extreme as you guys in Finland on kind of closing the shop completely in July. But now after having formed a Finnish gaming startup, I’ve learned the error of my ways and Iceland that my company now follows the finish July summer practice.

Joakim Achren 2:20
So you took the worst route us like, what are you guys like 14 founders and you’re taking holidays now the Finnish way?

Thor Gunnarson 2:30
Well, actually, it’s not because I didn’t agree to it because it was such a great idea to take the Finnish sort of Nordic style holiday, you know, year after founding the company, it’s more because we are working as fast as we can to stand our game up and get into closed alpha testing. So you know, as you know, as you know, running a game studio where you’ve got sort of staggered holidays happening over a two three month period, when you’ve got such a small team. It becomes super challenging to have, you know, all the discipline is present at any point in time when you’ve staggered the holidays. So at least this summer, we decided to just take our kind of allotted mandated holidays and take them all together. And it just just made sense to do it in July. I think next summer we’ll we’ll find ourselves with a larger team. And you know we’ll start doing a more conventional staggered kind of holiday approach as a company.

Joakim Achren 3:21
Yeah, makes sense. Hey, I want to get into your… first off your relationship to gaming and where you got started. But out of all the possibilities of careers and industries, how did you end up in gaming?

Thor Gunnarson 3:34
So I… my story is probably not dissimilar to many others in the industry. I grew up with games, starting at an experience I had when my mother was at medical studying to become a medical librarian at the University of Michigan in Ann Arbor, she used to take me along to the library and came across this computer terminal, turned out to be a mainframe computer and of course, the most popular application on that mainframe computer was a game called Spaceforce. That game had been around for probably 15 years at that point. But that was my first experience of a networked computer game. Then, of course, you know, grew up in video game arcades and then spent, you know, the 80s basically, buying personal computers. Vic 20, was my first computer, started programming in basic and thought I could become a computer programmer. But after a summer of programming, my k drive snarled, lost all of my work for that summer. And I don’t think I had the patience to become a computer programmer, but always loved the idea of being involved with the games industry. Took a kind of detour briefly into journalism and the music industry. But back in 1995, I get involved with sort of the first Icelandic tech startup that was doing something with direct relevance to video games. It was a startup called OZ Interactive. I was the first guy in the states as well. In California at the time founded the US subsidiary. And we were making basically some of the first multi user avatar based virtual worlds online. This is before we had, you know, 3d video games and networked video games. There was a really early attempt to create what people these days are talking about as the metaverse. Of course, we were probably a decade too early and way too young. We just didn’t know what we didn’t know. And sometimes that’s a great thing. But it was a great early experience. And then I went into the games industry proper. When I joined the mobile game studio in London, in 2000, called Ideaworks3D, we get a number of early 1.0 video games, did a lot of work on Nokia N-Gage that’s how I got my connection to Finland back in the day.

Thor Gunnarson 5:48
And from there, I did a series of mobile games for about seven years stint. Then join CCP in 2007 as the head of business development there and basically it got back with the A bunch of guys that had worked with OZ. Many of the guys who were at OZ back in the 90s went on to form CCP, created EVE Online, and I joined a few years after EVE shipped there. But you know, I think the games industry for me is just super interesting because it it kind of brings together heavy focus on technology, and technical innovation, but also creativity. And it’s sort of that combination, creativity and tech, that for me is just incredibly interesting. And as a guy coming from the business side of the equation, I feel like I just have the best job in the world sitting next to some of the most creative, insane game developers in the world and helping augment make their visions of success. And that’s really how I got into the industry.

Joakim Achren 6:45
Do you think like, from your perspective, technology, driving, ultimately, like new gaming experiences, would you say that CCP was a company that was built on that premise as well that you wanted to push stuff and Do you think that that is something that is seen now in the Icelandic culture as well that, you know, because in Finland, I think we have a culture of Let’s be really creative and a bit of like an artsy as well with the games in Finland. I think that’s how we see, at least I feel the Finnish culture is built up. That’s how Supercell I think acts as well that they want to, like craft the most good looking games, and then they figure out if it works at the last minute, you know, and they don’t really care about the well, the data is definitely there to validate. But it’s sort of like not driving, but it’s more creative. They think that the technology had a role in in CCP and Icelandic things?

Thor Gunnarson 7:43
Yeah, definitely. I think CCP, in its early days was a super interesting company because the founding team that came together on the one hand had that sort of creative, insane bent, which he kind of described initially in the demo scene. The man in CCPs case, one of the co-founders was Reynir Harðarson. He basically was the first guy, founding team there and he had the artistic kind of visual vision for the project. And I think one of his sort of favourite one liners at the time was “if looks good, it is good.” Of course, just the visual layer wouldn’t have been enough but really the guys that were on the programming team in the early days, people like in Hilmar was the current CEO of CCP, guys like Caspar Pierre, who came from a physics background with a PhD in physics and basically wrote large portions of bidding game physics engine and the in-game economy for EVE. They came from a strong programming background. And when EVE was created, there were no game engines, there was no multiplayer stack, much less a stack to create a single sharted MMO. So to get even close to enabling the creative vision I think the secret of CCP success was on the one hand, some really strong backgrounds in engineering and programming, but not from the video games industry because he was the first game created in Iceland. So all of the people that were on that initial team about 30 people in a ship EVE, nobody had made a video game at the time. This was back in 2003, when they shipped the first first beta of EVE. So yeah, in some ways, it’s similar to the Finnish culture in the demo scene, which is sort of some of the roots of the games industry in Finland, I find that the cultures are actually eerily similar between Finland and Iceland. And it’s actually one of the reasons why I felt so strongly about reaching out to my friends and family network in Finland when we were thinking about forming mainframes. Because the idea is large. It needed some pretty disparate skill sets. You know, whether it’s, you know, the MMO experience that we have from the team that needs to be a CCP, stretching to mobile game development and live operations to triple A PC, console, which is where some of my co-founders with their pedigree from Remedy certainly had that expertise. And so getting kind of a larger talent pool to create and enable division for Mainframe was critical. But it was also critical to do it in Finland from European because of the kind of unique experience that you guys have in the scene there. So that you know, the third part was the culture was super similar. I’ve been fortunate enough to have spent time in Finland, stretching back to the early 00s, a lot of good friends there and then sort of tech and startup scene. And I just knew that the culture fit would be super strong. And you know, I’ve been involved in so many companies where multinational workforce spread across multiple continents, creates, you know, all kinds of challenges in terms of you maintaining a common vision, particularly in the early days of a startup, where that kind of general cultural background is super helpful.

Joakim Achren 10:58
The first time that I have heard about you was last year from Samuli Syvähuoko, who is our mutual friend. And one of the investors in Mainframe, he was pointing out that this is super interesting company coming up now with looking at cloud based gaming. And I think that was… it was still very early in like the hype of the whole cloud based nature. But like going back to like you being at CCP then leaving, you guys had a startup, was it VR and then you transition to Mainframe. Can you talk about that story on how the cloud based gaming actually came up?

Thor Gunnarson 11:40
Yeah, absolutely. So I left CCP in late 2014 to become a co-founder in a VR studio in Reykjavik called Solfar Studios formed that company with Reynir, previous co-founder of CCP and Kjartan Emilsson who was the principal game designer at CCP at the time, we were super excited about the promise of VR and the prospect of VR deliver on some of the kind of fantasies that I think we all have as game developers about, you know, making SnowCrash real, creating this incredibly immersive virtual experience. And the other reason we started Solfar as we wanted to get back to a smaller team, we had all of us been working for a number of years and senior executive positions at CCP, you know, that’s easier to keep was 600 person studio stuff spread across four different studios, you know, crazy time zones, trip stretches, and so on. And we just, you know, we weren’t involved directly in creating games, any of u,s and we just wanted to get back to that. So so far was very much experienced in “Hey, can we use a modern engine like Unreal Engine, do you know create a very small team that actually where everyone is basically working in primary and secondary and sometimes tertiary roles in the development and you know, create some really VR projects to see if this was a medium.”

Thor Gunnarson 13:03
And of course the market that was interesting and compelling. Over the time Solfar, we found that the medium was super compelling. I think I’ve just had a blast working on projects like Everest VR and a roguelike shooter in-death, the two projects that we did at the time, so creatively, incredibly interesting, market wise, the growth of VR just wasn’t moving at a pace, where I think we had probably the patience to kind of wait around for the next wave of VR to really kind of grow the market and create the kind of space that we wanted to see because ultimately, our passion is in creating persistent online worlds. That was our pedigree coming from CCP and we saw an opportunity to create some some of the first network VR experiences.

Thor Gunnarson 13:52
Ultimately, I think we didn’t have the patient to patients to wait for that market to develop. And so the day that I started talking you In late 2018, probably about, you know, coming back to some of the ideas that we had been thinking about for many years about the type of persistent online world we wanted to create. And I had been following really closely, kind of really technical tests happening with gaming. Of course, we’d all seen the first wave of gaming back in 2013, 2014 period. But what I was seeing with some of the technical tests from Nvidia with their close beta of GeForce Now and project stream from Google, it was just obvious that you know, the underlying infrastructure like bandwidth capabilities to do the video encoding, and good technical quality of the stream was basically falling into place. And so the idea really was if we’re creating a world… if our idea kind of our passion is to create a persistent virtual world that can reach into people’s lives throughout the day. Cloud gaming engaged streaming is one of the technical innovations that allows us to create a world that people can access, where they can find a window to the world, whether they are playing on their mobile phone, on their laptop, or sitting in front of on the couch in front of your TV. And this idea of taking that genre of gaming the MMO basically, RPG, which is a genre of games that has some of the highest engagement and retention metrics in any type of game, to take that and supercharge that engagement and retention capability by providing a window across mobile PC console, or set top box was incredibly compelling to us because we knew from our past experience at CCP that one of the first studios to do a successful MMO on these emerging cloud gaming platforms, could have a real shot and creating kind of really the next generation virtual world. And that was sort of the technical kind of the understanding that the tech was coming that we would see probably most of the large platform companies and tech companies have offerings in this area was pretty obvious to us. And we thought, look, this is a technical innovation in the industry that comes along very seldom. We have a concept that we think is a perfect match for this upcoming technical innovation. And let’s get started now.

Thor Gunnarson 16:29
So what we actually did is we essentially started flying out to Helsinki and spent time sitting at a pub somewhere in coffee was the first time I discussed our idea with anyone and Samuli [Syvähuoko] coined a term that I don’t think I’ve used in public before but he said, Thor this is crazy. This is quadruple A gaming. You know what you have in your hands here as excitedly don’t say that to anyone but I’ll say it here now because it was just such a passionate kind of enthusiastic response to the concept, because of course, we were aiming to create a game that had the immersive qualities of the triple A production, but available to any screen to your phone. Know the kinds of gaming experience you simply can’t see mobile gaming today. That was kind of our idea. So so with Solfar, Samuli and Sisu [Game Ventures], were investors in Solfar, they were part of the early angel round in Sofar, he encouraged me to like, “Okay, look, can you figure out a way to get it, what you guys are doing currently… Why don’t you start working on some prototyping around this and some ideas,” and so during the winter, from 2018 to 2019, we started doing 2% prototype, visual target of what this idea might be, presented it to both you know, prospective investors and most importantly the people that would go on to forming from within Helsinki with us. And so, you know, I spent the Christmas in Iceland with Saku Lehtinen and who is our creative director and was previously a Remedy from the beginning. He actually gave some incredibly useful kind of feedback. And thoughts on the concept. And the concept kind of changed, I think quite helpfully and dramatically as he got involved with it. And then as we went through the winter, we started, you know, reaching out to people in our network getting introduced to developers in Helsinki. And we found that we had a concept that was both resonating with a bunch of people, but also was a really exciting prospect for I think the Finnish dev scene, or the developer scene, because in Finland, you guys have on the one hand, an amazing mobile game scene, you have your handful of really good triple A, PC and console studios. But historically you haven’t created… you certainly haven’t created an MMO, the closest to that was Habbo Hotel, which was the closest kind of match to an MMO RPG, I think historically. So we found a lot of people that were super excited about the concept working on a different type of game that leaned on their previous expertise from mobile gaming, AAA and so on. We, you know, kind of went through the winter in forming a prospective team. And ultimately we founded Mainframe in Helsinki, because we felt that the appropriate home for our company was Finland, rather than Iceland in terms of the headquarters of the company. And along the way, we basically acquired Solfar and folded into Mainframe, such that so far basically became an operating subsidiary, we shifted out of the VR development and took kind of the prototype that we’ve created so far and rolled it into the newly formed company in Helsinki. So in one sense, the the story is sort of a pivot story. But I… you know, I think of it as something much larger because the concept itself would not really have taken off unless we had found these amazing co-founders in Finland to join and kind of you know, take the concept further forward.

Joakim Achren 19:51
Yeah. Thank you guys haven’t really talked about the game much publicly. What it’s about but can… like this is I think, One of the most interesting things about cloud gaming is like, if you could share any thoughts on how game design can actually take advantage of cloud gaming, what are the things that Mainframe is going after at the moment?

Thor Gunnarson 20:14
Absolutely. So yeah, you’re right, we haven’t shared much about the game, the specific setting without really any details beyond to say that we are creating a sandbox MMO game that has a very heavy emphasis on the social layer and the set putting, creating the setting and the game environment that allows a social network to form within the game, which is sort of a feature of sandbox MMOs that we saw with EVE Online, certainly over the years, and apply that concept to a cloud gaming context where we have again, this ability to have players to have a touch point with each other throughout the day, irrespective of you know, which screen they’re on any given time. So this idea of sort of supercharging social fabric and social connectivity in the game, throughout the day to us was incredibly compelling. So that that’s one part of it. That’s one aspect of why we believe a virtual world that you can access throughout the day on any screen is so powerful because it really, really strongly emphasise the social aspect of your game experience. So you can have your short, rather casual sessions when you’re on the bus heading into work. But you know, you might then find yourself at home, whether it’s on your laptop or your TV than playing a longer session. But it’s the same game. It’s the same visual experience. It’s not kind of a watered down side application of the main experience. It’s the same game as all of these fields. That was sort of the key facet that we started to focus on.

Thor Gunnarson 21:44
But when we think about what cloud gaming brings, I think more generically to game developers. On the one hand, you know, of course, I think everyone is aware now of the prospect of frictionless engagement and user acquisition with the game. This idea of seeing a person playing the game on Twitch or on YouTube or Facebook gaming, and receiving an immediate drop, allowing you to jump into the game, join the game is incredibly compelling from a acquisition perspective and from an engagement perspective, and leads to what we think is incredibly important is this notion of social onboarding. We used to call this virality back in the Facebook, early Facebook games and social game era. But there’s something incredibly powerful about a personal recommendation, referral from another person, your friend or someone that you respect or a following whether it’s an influencer or a streamer on Twitch or a friend of yours. And being able to jump into that experience and be on boarded into the game from the gecko by a person that is playing the game. We just think that’s incredibly powerful. And cloud gaming and streaming has the prospect of allowing that kind of dynamic to happen. So that’s kind of more on the front end and on how the funnel might change with plugins. But from a developer perspective, also, there are some really interesting kind of technical facets of gaming that can be applied to game design. So I’ll take just a couple of examples. In any online game, cheating and hacking as a massive problem, the fact that you put out a map and it’s instantly known to all players, because the data is basically stored on a PC, I can easily reverse engineer, so leads to a lot leads to wall hacking, it leads to all of the problems that you see companies like Valve, you know, just to give an example of CS Go, fighting, constant hacking and cheating that happens again, and ruins the experience in many ways. For many, particularly more casual and new core players. In the MMO space, this has been a huge issue for us over the over the years. With a game that is delivered as effectively a video frame to your device, the client is never present. The data is never present on the player device. It’s streamed from network and streamed from the cloud. It becomes very hard to hack the game plans never say impossible because hackers you know, obviously have no tenacity and all kinds of interesting ways of getting at data. But generally, by and large because the client runs in the cloud, it’s you can you can trust the client. So that leads to all kinds of super interesting game design and level design capabilities that we simply haven’t had before. So if you create a large, for example, large open world where the map is mystery, where it is incredibly hard to know the map or be able to spy the map, it allows you to do just incredible stuff from a level design, from a world set design perspective, and ultimately leads to you know, that ability to trust the client leads to an online experience that can be trusted becomes more approachable, accessible to players generally, because you know, we’ve all had this happen to us you jump into first person shooter, somebody is running and in-game bot, the client’s been hacked. And it’s a horrible experience for a first time player.

Thor Gunnarson 25:02
You know, there are just very relatively few of us that have the patience to, speaking generally about gamers, that patients to kind of get engaged in the game that happens suffers from that sort of problem. So I think that, you know, I will have, you know, obviously examples of this when we say more about the concept of the game, and so on. But that that’s one example. The other thing that’s really compelling from online games perspective is that because the client and the server are running in the same network, typically, you are able to take advantage of some very fast networking between the client and server running in the cloud. So consequently, you have lower latency, and the ability to ideally, have larger player populations in your map, which for an MMO is incredibly compelling ability. So you’re not dealing with the traditional client server issues that an online game has to tackle, you know, be also from a game developer perspective. Get the best benefit of targeting a single runtime in the cloud, more or less. Whereas you typically a game development consists of doing multi platform development project, where you’re spending a lot of time importing and QA to get the game to work even across the minimum spec of PC devices, right? So being able to target a much larger audience with less of a porting and QA effort is, you know, just logistically a huge benefit. It allows us to spend more time and effort on the game itself, then on the technical kind of facets of porting and compatibility testing, and so on, so forth. So that’s a couple of examples. There’s a bunch more, but those, those are a handful that jumped out at us. It was Yeah,

Joakim Achren 26:45
Those are definitely interesting. Do you hear about like a challenge, usually for a startup is the product market fit phase? Can you share your thoughts on how do you actually see a cloud gaming game reached product market fit like it? What kind of validation are you guys seeking at the moment for your game?

Thor Gunnarson 27:09
So I think there are two critical moments for us in the near future. On the one hand, like any MMO developer, we need to get the game in front of our players as quickly as possible. So we are working very fast to get to that closed alpha moments where we’re able to test our core loops, we’re able to bring the first few hundred first few thousand players to the game very early on, to judge and sort of understand and follow what they’re actually doing in that test. Because it’s sort of the hallmark of the sandbox MMO, which is the category that we’re focused on that players will do very unexpected, kind of sometimes illogical things in your world, and you sort of have to follow the hierarchies and structures that begin to work. That’s one critical moment for us in terms of product more Fit. The second one, which is specific to cloud gaming is the UI/UX layer and the user experience in the game. That challenge for us is a non-trivial design challenge because we are aiming to create a game that treats mobile as an equal citizen to the PC or console experience that delivers experience to mobile devices and phones that feels native to the platform that is accessible and can support the kind of shorter gameplay sessions that you would see on mobile typically, and allows players to feel that they they can have a satisfying experience again, which is not sort of a second class experience to what they see on PC or console. And, you know, that’s that’s a challenging design exercise. We’ve made some really positive progress. We’re continuing to build out a team comprising UI/UX designers with backgrounds in mobile, PC, and kind of bringing that all together, hopefully to deliver a user experience that really resonates with people on mobile on. So that’s a challenge, we need to get to a closed alpha test to get that user testing done really early on in the development cycle, because it’s sort of one of the key challenges and opportunities of doing a game that is playable across mobile and larger screens.

Joakim Achren 29:21
Yeah, definitely. Like, the whole idea about cloud gaming. I think everybody’s been talking about the barrier, which is it’s the server costs that you might accumulate this your… You’re basically taking user commands and processing them and then the frames versus where Netflix can just no use the CDN to send out the same material to everybody. Like I presume you worked out some LTV calculations to actually achieve profitability. Like for a cloud game, like can you talk about the cost structure of running a cloud game like what you guys would be doing on an AWS or something? Or do you need something more specific to achieve it?

Thor Gunnarson 30:04
Yeah, it’s a great question. So I think the way I think about this is in sort of many ways, similar to what we’ve seen in any kind of major technology or platform shift in the games industry. Yeah, you know, when we look back at the the advent of 3d acceleration in the 90s, you know, the first 3d effects, GPUs were quite expensive, far outside of the mainstream. And you know, nobody was really doing 3d gaming at the time because it was required some some expense and so on, so forth. Same thing happened with the shift to online gaming. The slow rollout of broadband capability took a number of years going stretching from first experiences on a 44 k dial up modem to, you know, a broad kind of swathe of people that actually had decent broadband access. So that took some time and you know, it was definitely kind of a non mainstream experience early on.

Thor Gunnarson 30:58
The second part of it kind of thinking about cloud gaming and the economics of cloud gaming is that this is a scale exercise. This is, you know, kind of a game that, of course, requires the scale and investment ability of the large tech companies that are entering the market now. So the likes of Google, of course, Stadia, Microsoft, xCloud, Tencent, and Nvidia, are the companies that are building out the infrastructure that brings down the operating expenditure. And really, that’s measured in minutes a stream or hours of streaming. And there’s a certain number that you have to reach to make the whole exercise economical. And I think we can assume that the folks at these companies have done the math, they’ve worked out that there is an economy of scale here, it’ll get less expensive for them over time as more people come to these services. And in that sense, it’s very similar to the cost of building out for example, a mobile network makes no sense to build that mobile network for one person, but when you’re building it for millions of people, those economies of scale and net effects become important. So I think it’s challenging if you’re going to run directly on to kind of existing cloud networks and spin up an Nvidia instance in the cloud, those cost structures are challenging. But I think if you’re viewing it as I’m a game developer coming to a platform like Google Stadia like a project xCloud, you are assuming that the cost structure is something that are effectively being amortized, or carried by the platform rather than you covering the permanent cost of that stream.

Joakim Achren 32:31
Yeah, just like I want to dive deeper into this one and be like there’s, there are other startups who are coming into Cloud gaming at the moment. And they’re thinking about this as well as the biggest barrier. But what is your advice for startup like founder who doesn’t have all the expertise the network who’s thinking about cloud gaming as a viable thing to go after at the moment? What would you advise them regarding the unit economics cloud gaming?

Thor Gunnarson 33:01
So if you’re a game studio and you’re creating content for cloud gaming or game that is designed to be cloud native, you do have to care about the unit economics and the operating expenditure of the game, you probably are designing the game somewhat differently to support more frequent, shorter gameplay sessions. So there are some factors to think about around that. But, you know, again, I think if cloud gaming takes off at all, the unit economics will scale because of the presence of the tech companies and the tech giants that are basically running the underlying infrastructure. So I would actually say, yes, if you took the rate card from Azure or AWS today and tried to do the skip the math, the math doesn’t look that great on a standalone basis. That’s absolutely is true. But again, we assume that you know, economies is a scale will kick in and that the infrastructure in the data centres that companies like AWS, Google and Microsoft have around the world, obviates the need for a game developer care as much about this unit economics compared to if he were he or she were basically running the full technical infrastructure themselves.

Thor Gunnarson 34:11
You know, look, it’s the same. Let’s think about this way. If you wanted to create a console game, the first thing you would not do is create the game console. file. The cloud network is the game console. It is all of the hardware, the bandwidth infrastructure that is needed to enable a creative vision of a game. It requires Sony and Microsoft to invest substantially in the early days and at the launch of each console cycle, to create the underlying ecosystem and the platforms on top of which game developers realise their vision. And so, in the same way that we saw Sony innovate the business model in the 90s, basically following sort of Gillette Razor and Razor blade model with their first PlayStation, they took a loss on the hardware. I made it up in software licencing and software sales on top of the platform then made it up as they drove down the cost of goods sold for the PlayStation to start turning a margin on the hardware itself. I would think about cloud gaming exactly the same way. We rely upon companies to innovate to create the tech platform upon which our games run. So so that’s kind of how, certainly if your game studio you kind of as the metaphor that you should really be considering, think about doing something on cloud gaming.

Joakim Achren 35:24
Yeah. One more on the whole cloud infrastructure thought like, because in mobile, if you just go to what’s happening on, you know, the casual, download mobile games industry, you have players, like Playfab, Gamesparks, which are tools for creating server infrastructure, basically, like out of the box. Do you think that, you know, there needs to be more games out there before these kind of startups emerge? We’re bringing the middleware or the tools needed, the shovels to actually get more companies into cloud gaming, or do you think it’s up to Microsoft and Sony? What do you think?

Thor Gunnarson 36:09
Yeah, that’s a great question. So on one level, creating, creating a game for the cloud is no different than any other platform, you typically will choose a Unity or Unreal. In our case, we’re big fans of unreal and on top of Unreal, you are really relying upon the cloud platforms to provide you with the streaming capability. And you know, most likely some sort of thin client or access to the browser on the player side. It when it comes to online gaming and networking, yeah, I mean, there are a number of kind of middleware, companies and services out there. They tend to find opportunities where they can fill feature gaps that are not being covered by unity or epic. middleware is an absolutely critical part of any modern game studio. Most of us rely upon the large existing engines like Unreal, Unity. We’re fans of Unreal Engine for the type of game we’re creating. But both are very compelling. And there are opportunities for middleware companies to create sort of discrete parts of the stack that are not being served adequately by Unity or Epic. Online gaming has been one of those aspects. So yeah, I think there are opportunities for middleware on the online or networking side, I haven’t seen yet any kind of pure examples of middleware created specifically for cloud gaming. So that may come further down the line. But at the moment, in our studio, we have a very specific style of game that we’re creating. It’s a large scale MMO. So we are doing a little bit of internal tech on our site to enable that on top of Unreal Engine.

Joakim Achren 37:47
Yeah, it definitely sounds like something that will come. As long as people are coming into the industry trying things out, there will be the middleware as well coming in, just gradually going to happen.

Thor Gunnarson 37:59
I think that’s true. I do think that, you know, the challenge with a middleware startup is in the games industry is that Unreal and Unity kind of suck a lot of the oxygen out of the market, that the two companies have become so dominant that you sometimes have a opportunity to create a specific capability that then you sell to Epic or to Unity. It’s kind of hard to create a standalone super successful middleware company these days just because of the scale and stature of these two plans.

Joakim Achren 38:31
Yeah. How would I move on to discussing fundraising? Like, how did you learn to do fundraising for so far, and then successfully do that again, from what mainframe? Like? What was the process for you,

Thor Gunnarson 38:45
Uh, specifically with Solfar and then with Mainframe, it was very much about meeting on our existing network, reaching out to people that we’ve had some past experience with and obviously bringing a story, an initial team and a thesis for the company that was compelling to those investors. So you know, experience and your network is part of that. If I speak frankly about where I learned kind of how I approached fundraising Initially, I learned it initially by making a horrible mistake. The first time I went to an investor, I was at oz interactive, virtual worlds company I talked about in the mid 90s. We were trying to raise money for our company at the time, went up to Silicon Valley. And those days, you know, we didn’t have PowerPoint. So we created a business plan. I was about 100 pages, wire bound with pretty pictures and took it to investors on Sand Hill Road in Silicon Valley. And, you know, we had no idea what we were doing, but and we failed to raise money at that time. But we took learning lessons from it that we felt were just, you know, a learning experience. And you have to take that kind of perspective when you’re hitting the road and starting to fundraising for the first time. We subsequently raised successful for the company and as you know, multiple rounds of successful funding. Specifically, when it comes to software, our main thing, my approach, there was both cases to move pretty rapidly to raise, raise, you know, whether it’s an angel round or very early seed round, get it done with investors that you ideally know or are getting strong recommendation from others that you know, and try to look for a match with those people because they are, you know, they’re coming to your company very early on, and they’re going to be with your company until the end, you know, so getting that that kind of match and that that that personality fit is incredibly important. And also the early investors that you really lean upon when it comes to obviously raising additional funding, expanding your network getting credibility as a startup. And I think, you know, certainly in the case of mainframe on the seed round that we raised, principally in the Nordics was a really good example of that, because we brought a really good cross section of developers, stripe investors, from Finland, investors here in Iceland. And they gave us that kind of local network, that local credibility that the viewer can then use to kind of step up to the next level and reach out to a broader base of investors.

Joakim Achren 41:26
Yeah. So in a sense, like, you guys had something that was compelling. Like, if that would have been the case, would you even have started discussions with it? Was it like, because I think at that stage, like, you guys knew that this was compelling that will go after funding. You didn’t need more convincing, right?

Thor Gunnarson 41:49
I think we we you always have to, you have to show some form of traction. In our case, we did some very early prototyping that we were able to do. Already in the kind of first first round of funding this you know the story and how you present it is critical, you have to have some form of thesis about the macro that you’re seeking investment for. In our case, I think our understanding and take on the prospects of cloud gaming, I think was compelling to investors and probably made our story more compelling than if we had simply brought an MMO concept and studio to our investors. So, of course, when you’re looking to form a game studio with venture backing, you’re looking for some defensible advantage in what you’re creating, you’re looking at, ideally, bringing a company to market that has a finding advantage. So first mover advantage in an emerging technology space is often one way to do it. And so I think certainly when we were presenting mainframe was the combination of very deeply experienced team across different corners of the industry. A strong creative story for the project based on past experience and track record. And then finally, you know, very strong thesis about a potential first mover advantage for the

Joakim Achren 43:13
Yeah. Can you talk more about the conviction that the VCs had for your Series A round? You’ve got Andreessen, Riot participated. What was the conviction that they got on the timing, that now is the right time to jump into cloud gaming?

Thor Gunnarson 43:31
So I think the conviction and timing question is part you know, you’re looking to talk, to reach out to investors that are demonstrating an understanding of the category and understanding of the opportunity, irrespective of your pitch, if that makes any sense. So they typically have formed their own viewpoint and their own thesis about the opportunity and are looking for companies that exemplify some of those, some of those trends and you know, when you’re looking to reach out to investors, you know, you, of course are doing your homework, getting recommendations and trying to understand what they’ve invested in the category previously. And I think in the case of Andreessen Horowitz, they are doing increasingly big investments in the games sector, we were aware of them and one of my co-founders Sulka Haro had a personal relationship from the past days with Andrew Chen, partner a16z. And so we were able to set up a very quick initial conversation, it became evident in that first half hour call that there was a strong match in terms of how we were viewing the industry and the opportunity and how they were had come to their own conclusions about the future of gaming. So it’s timing and homework, it’s making sure that you don’t spend a tonne of time speaking to potential investors that do not have thesis in this category and do not have a track record in the category. And you know, if you get right to you kind of find investors at the moment that they are forming, you know, thoughtful takes on the space and the opportunity. And if you’re there at the right moment in time, getting them to conviction about the opportunity, because a lot easier.

Joakim Achren 45:11
I think the whole pointing out that this is our belief where the industry is going from the VC side, that would be awesome. If that would be mandated for every VC operating and gaming that it should think about, like what they’re looking at going after.

Thor Gunnarson 45:28
Yeah, absolutely.

Joakim Achren 45:30
it didn’t end up ends up in a situation where, you know, you need to pitch everybody to understand who wants you doesn’t?

Thor Gunnarson 45:38
Yeah, that’s absolutely right. And I think a16z they exemplify, you know, kind of the gold standard of how you communicate to the wider ecosystem and community. The a16z podcast and blog are places where the partners are, you know, putting some public statements down and really useful information often as well. And certainly The series of blogs that john Lyon Andrew authored, in the in that window as we were fundraising, it just became super evident that there was a really strong match from from, from the perspective of condition condition and the thesis about the industry.

Joakim Achren 46:15
Question about the round like you raised a pretty big, nice $8 million round. How do you believe in this kind of like balancing your cash flow with the expectation for a product? And also updating your investors on the milestones? Like, do you preemptively think about having a management team in place who can see where things are going? How do you structure kind of like the visibility into like, what’s the runway looking like? Are you hitting your milestones and things like that?

Thor Gunnarson 46:47
Yeah, I mean, we think constantly about it, you raise on the basis of a milestone that you can reach and traction, you can demonstrate in that net runway and in that time that you have particularly when you when you Are when you’re in an industry that is, is sort of an emerging corner of the games industry. Being in close contact with your investors and with your board is absolutely critical. So we try to update and communicate as much of our progress and we set ourselves some very clear milestones that the use of proceeds from the investment is quite clear to our investors. So you have to very carefully put that together and your operating plan has to have some clear, achievable milestones along the way. And hopefully, we’ll be able to exemplify that as we as we proceed.

Joakim Achren 47:40
Yeah, that’s awesome. Then going back to the Icelandic gaming culture, I still wanted to touch base on the entrepreneurship side. You guys have, I think, the most successful startups in the world where per capita, like why do you think that has formed up in Iceland? The whole entrepreneurship culture?

Thor Gunnarson 48:02
Yeah, that’s a great question. Um, so it probably starts with a couple of factors. We have a that’s not dissimilar from the rest of the Nordics. We have a very edgy, highly educated, literate culture. That kind of underlying culture of Iceland is is one that has quite a bit of respect for creativity. It’s often said that every Icelander has one book in their stomach, and they’re just waiting for the right moment to write it. Your music scene here in Iceland is another example of that creativity. We just have an incredibly vibrant music scene that has been ongoing now for the last 25 years. You should we shouldn’t be able to have this strong music scene in such a small country. I think that the other factor that is incredibly important and that relates to the startup scene in Iceland is there has been a high tolerance for risk taking in Icelandic culture. people you know, there’s there is no shame in going out and forming a band or telling someone that you’re about to go and write a book for a year, or indeed form a games company. People support that sort of culture here of doing something and failing at it and support it. I think that’s important. The other factor for perhaps why we’ve seen a pretty vibrant games industry in Iceland, given our small size is of course, we have a cornerstone company here in the form of CCP games, founded in Iceland, who to become a very successful studio, yours with mine, of course. And anytime you have a cornerstone company in a particular country or city, those companies they tend to be forum ecosystem of startups around them. Typically where you see team members or founders or executives leave that initial company and formed their own companies. And you start to see then a ecosystem of smaller companies. These are typically coming out of people that that were part of that cornerstone team. So you guys would have a similar situation with Rovio or Supercell, in Helsinki, or Remedy. In Stockholm, it’s DICE. You could go on and on, you can kind of you can pick the city. And you’ll always find kind of an initial, you know, Ground Zero studio that got that initial success. And that led then to kind of influencing and supporting the underlying system.

Joakim Achren 50:29
Yeah, that is so true. Back then, then you also if you have a culture where people aren’t conforming to do what everybody wants to do that conformity is that everybody has to have their own thing. That’s like super awesome. I think that’s what’s going on in at least in this Nordic countries.

Thor Gunnarson 50:49
I think so I think it’s I see it all the time. When I’m in Helsinki. I mean, just the startup culture in Finland is so vibrant, then just amazing to see how that’s built. From when I was first spending time in Helsinki working with Nokia back in the early aughts, it’s just been an amazing ride over the last few years.

Joakim Achren 51:09
Hey, I have some final questions for you Thor. The The first one is, so I was recently writing about pre-mortems, which is kind of like a system for coming up with the ways that a project could fail before the actual project starts. Sort of, like preemptively thinking about, “Hey, this is how we’re going to fail. And then we build, build based on the learnings before it actually happens.” Like, do you have any words… have you scouted kind of the way Mainframe could failed, you do end up screaming at middle of the night because you’re thinking about those things. How do you see that kind of approach?

Thor Gunnarson 51:47
Oh, of course. I mean, no founder in a startup sleeps peacefully at night. There are so many things that can go wrong. I love the idea of a pre-mortem and thinking about the points of failure upfront. I hadn’t thought exactly in those terms, I’ll probably dig into some of your writing about that, because it seems like an awesome way of thinking about it. But, you know, certainly, if you have some experience in your respective industry, and certainly our founding team had experience in game development, you know, we were very aware of the initial risks of just the team formation, how the team would actually work together, how, you know, kind of the fit individually would be, particularly because we took the path, which is quite unusual and starting up with a large co-founding team.

Thor Gunnarson 52:36
So we have 13 co-founders when we started Mainframe. Yeah, we did that partly because we wanted to de-risk the whole kind of build up team formation stack post forming of the company. So we actually did a tonne of work in the winter of 2019. Before we incorporated mainframe in reaching out and talking with prospective team members and co-founders, you know, Samuli and the guys at Sisu, who were instrumental in helping With that, during that winter before we formed an integral investment, and so yeah, I mean, the team dynamics is a critical point that every startup should be thinking about, it’s hard enough to make sure that the dynamics are good with two co-founders or three co-founders, which is the more common setup. In our case, we wanted to be able to hit the ground running and quite rapidly through our concept and prototyping stage as a company. So we try to de-risk the team formation team kind of risk by doing a lot of that homework and really work quite early on in the process, even before we formed the company. But yeah, we think about operationally all kinds of points of perspective failure. You know, I think the… when you’re trying to create a game that is some sense of creative and practical leap of faith, you of course, constantly think about, you know, did I get the timing right? Are we going to be bringing our project to market at the rights sweet spot, we’re… are you launching into a growing market? You know that that’s always a concern. But if you launch a startup with everything fully de-risked, then you’re probably not doing it. You know, there are always unknowns. That’s kind of just part and parcel of being a startup game studio.

Joakim Achren 54:18
Hey, can you name your favourite book? And why?

Thor Gunnarson 54:23
So I’ll give a fictional… the fiction example I’ll give is these days becoming a pretty standard one. So I was absolutely blown away by Neil Stevenson’s Snow Crash, when it first came out. If somebody was joking these days that you’re good to any executive in Silicon Valley, you’ll find a well-found copy of Snow Crash. That’s that’s probably true. I actually was inspired by it. I mean, a huge science fiction and fantasy. And but I was inspired by it, as were many of my colleagues at OZ back in the mid 90s. To the extent that you know, this idea of metaverse of this virtual world this virtual reality was so Strong that when we were creating our first multi user 3d worlds, at OZ we named the first world that we created the street as homage to Snow Crash. Yeah, so you know is a vision for the future that any good sci fi fan or game fan is going to love. And you know Neil Stevenson obviously after snow press wrote so many amazing books, Brooke cycle can go on the nonfiction example these days. I’m reading Ben Horowitz, his first book, The hard Thing About Hard Things. Then, of course, is one of the founders at a16z the story of his kind of early days at Netscape with Marc Andreessen. And then, as they founded LoudCloud and ultimately sold the company is just an amazing blow by blow account of all of the challenges and all of the grey areas that you will encounter if you are involved in a startup. It’s just it just resonates on so many levels to take away that there are no hard and fast, correct answers. It’s just incredibly compelling. So I recommend up to that book to anyone that start embarking on a start.

Joakim Achren 56:09
Yeah, that book is sort of like you’re in a boxing match, and you’ll get like punch, like, really hard immediately, and you need to survive.

Thor Gunnarson 56:18
Yeah, it’s such a great memoir of startup life. And you know, they’re just some great takeaways, I highly recommend it.

Joakim Achren 56:27
Do you have a story that shaped you and how you approach your work today.

Thor Gunnarson 56:32
So there are many stories, I suppose one that I sort of think back on is my first high stakes presentation and demonstration in the tech industry, mentioned the company I was involved with in the 90s. I was interactive. I was the only employee of the company at the time living in Los Angeles. The rest of the team was based in Iceland with a team of 15 people at that time. We had done our first trade show in San Jose at Internet World, which ones made trade show at the time and shown a version of our multi user avatar based world as a demo at that show, with, you know, a little bit of kind of fake it until you make it approach, which is that we showed the graphical environment and kind of the vision of the project, but we hadn’t actually finished the multi user code and our marketing and our pitch was kind of centred around the fact that we were bringing multi user 3d worlds to the internet. So we get a approached from a pretty senior business development by Intel, who bites us to come up to Santa Clara Intel’s headquarters and present the project to Andy Grove, who was at the time the CEO of Intel, and probably one of the more famous CEOs in Silicon Valley, historically.

Thor Gunnarson 57:54
I was down in Los Angeles, and our CTO at the time at Oz, Kjartan, who is now the CTO of Mainframe today, he was furiously at work in Iceland compiling and writing the code and debugging the code before this demo because he wasn’t ready. I was sitting in Los Angeles with a desktop machine downloading his patches and his updates over a modem waiting for the bill actually stopped working. My flight was about to leave, I was getting to a point where I was not going to be able to make my flight and get the meeting up in Santa Clara unless this code actually worked, got the final patch and didn’t have time to test it flew off straight from the airport with my desktop with the only working build of the project connected to the internet. First thing I saw when I booted up the application was the first avatar standing on the head pin in the virtual world. And so the takeaway there was no sometimes just circumstances are such that you have to take the risk, you kind of have to go for it. If the opposite Is there we were probably also young at the time that we didn’t know how much of a massive fuckup this would be if we showed up without this code working. But sometimes you have the opportunity presents itself and you have to grasp it. And probably that taught me some bad habits about risk taking and kind of engaging in high stakes adrenaline fueled presentations. But but it’s it kind of the takeaway I had was, you know, sometimes the time and the moment is there, and you have to take the risk, see if you can get there. And in our case, the demo worked. Andy Grove, walked into the room half an hour later, saw the demo. And within a month, our company was on stage with Andy Grove at the company Sun Valley Conference, which is sort of the most High Stakes kind of tech conference in the States. And that led to, frankly, our Series A round as a company and the company was off to the races.

Joakim Achren 59:53
That’s an awesome story. As the final question, what’s the best way for up and coming entrepreneurs, people want to ask you some questions. What’s the best way to get in contact with you?

Thor Gunnarson 1:00:05
Hit me up on Twitter. I’m at Thor Gunnarsson on Twitter. And you know, happy to take some DMs and questions there.

Joakim Achren 1:00:12
Well, hey, thanks so much for this interview. And best of luck with the game and I’m waiting to play.

Thor Gunnarson 1:00:20
Thanks Joakim. This was a real pleasure and hope to see you in Helsinki for too long.