"Our Mission is to Simplify Everything": Xsolla On Its Approach to Solving Cloud Gaming Issues

Xsolla's VP of Product Nick Bondarenko offered a comprehensive overview of the most common problems of the cloud gaming industry and explained how the company solves them by prioritizing partnership over competition.

In 2024, cloud gaming no longer seems like something exceptional. Last year, at GDC 2024, when we first announced XCG, it seemed like just a nice addition. But today, I can assert that in some cases, it's almost an irreplaceable tool for testing, porting, and providing demo and trial versions.

The reason lies in the cloud being ideal for game testing: from the "friends & family" phase to alpha and closed tests. It solves equipment compatibility issues, makes controlling access easy, and eliminates the risk of the project getting leaked. Examples like Destiny and EVE Online show how early versions of games running on the cloud can strengthen communities and provide valuable feedback in a fully controlled environment.

Cloud also solves the problem of high hardware requirements and porting difficulties. Games on Unreal Engine 5.3 with RTX, which are hard to port on mobile devices or devices with outdated hardware, are easily accessible through the cloud. Various websites launch mobile games in the browser, Jam breathes life into old PC games, and Care Games allows playing P2P mobile games on PC.

Launching a mobile game in the browser deserves a special mention. The cloud offers the opportunity to test your metrics for the web audience in just a few hours with zero investment in R&D.

From the list of the previous year's hypotheses, we found that the "try before you buy" and "play before downloading" use cases can be considered successful. Especially for games worth $35+ at launch without a discount.

Sometimes, 3-5-year-old games just can't be ported. Studios are busy or closed, and investments in a porting house are too high. And the cloud is the perfect solution to give such games a new life on a new platform. In a custdev interview, the CEO of one sci-fi MMO noted that most GFN players choose the free service and play for up to an hour, which does not require a payment. However, data shows that about 25% of users prefer the pay-as-you-go approach and play comfortably for 5-8 hours a week. In general, cloud gaming in 2024 is about accessibility and coverage, which can be extremely difficult to achieve through traditional porting to a new platform.

A recent report shows that with a significant 91% ownership of PCs, as well as 86% ownership of Androids and iPhones, a majority of gamers have the necessary hardware and smartphone compatibility to support cloud gaming.

More than 56% of respondents appreciated the opportunity to try a game before committing to a full download, purchase, or subscription, emphasizing how important gamers find previewing their games and informed decision-making.

What are the current problems of the cloud?

At first glance, it seems that cloud gaming is simple if you don't need to focus on the streaming technology itself. But as we began to understand how everything works, we often recalled the "one does not simply..." meme.

I would say that it all starts with understanding that in cloud gaming, each region is a separate launch, not like in Steam, Epic, or Google Store, where the availability of the game in different countries is regulated with a few clicks.

The complexity begins with the non-obviousness of pricing. You see, the cost of an hour/minute of streaming, the cost of storing your game, the cost of streaming. And then the unique combination of the factors for each region or data center where you place your game. You'd like to think, "I'll release my game, and one hour of gameplay will cost $0.5". In reality, it's going to be "I'll release my game and 1 hour will cost between $0.4 and $0.8, depending on where the player is from".

Cloud gaming is often advertised as an opportunity to launch a game quickly, but in reality, to ensure immediate access, you need to have many servers ready for instant launch, which significantly increases costs. A "cold" launch can take from 1 to 5 minutes, depending on the vendor. To reduce this time, publishers are offered to reserve servers, guaranteeing faster access for a certain number of CCUs.

For example, if you have reserved servers for 5 CCUs, they will be able to launch the game in 8-20 seconds, but the 6th player will already face a delay. This creates a direct dependency between the speed of access to the game and the amount of reserved server capacity, requiring additional investments for each region separately. Want to quickly enter your game in 5 different regions of the world? Yes, you got it right – you need a reservation in all 5 regions!

The lack of MASG (Multi-Application Streaming Groups) support complicates reservation further. Without MASG, you need 25 servers for 5 games in 5 regions. MASG allows one server to support multiple games, like having 5 laptops with different games at a party, but only 5 people can play at once.

You've probably seen forecasts for the growth of cloud gaming based on the ever-increasing internet speed around the world. It's true, but if you live a couple of thousand miles from the data center, the quality of streaming will begin to degrade. Do you know why it's not very convenient to play from Egypt, Italy, or Argentina in 2024? Most of the available regions for cloud servers coincide with the location of major data centers, leaving the "last mile" problem as a key challenge for cloud gaming.

How do we solve problems?

I believe that you, as a publisher, should not pay for unplayed games. There's no need to delve into the details of reservation and how it all works. Adding new regions should be as simple as in Google Store – just click a checkbox next to the country. I want your experience with cloud gaming to be as close as possible to traditional platforms, without the need for special porting of games in 2024.

Our mission is to simplify everything.

It's no secret that at Xsolla, we always prioritize partnership over competition, seeing this as our strength. In the field of cloud gaming, our approach is to aggregate cloud technology and hardware vendors to provide game developers and publishers with maximum coverage and a universal solution for the quickest launch of their games. If one vendor excels in North America and another offers better prices in Asia, we choose the best option for each market. For developers, this process is transparent and simplified, from deployment to scaling and load management. Thanks to a single API, our partners can manage the distribution of their games in real-time, minimizing technical and administrative complexities.

Aggregation allows us to reduce costs and speed up the launch of games. We are working on creating cross-game MASGs, optimizing the use of equipment for the maximum number of games depending on the current CCUs and peak loads. This allows us to efficiently distribute resources in real-time, avoiding unnecessary expenses for reservation. This is very important, as it allows not to pay for reservation in case of peak loads.

Non-obvious ways to reduce equipment costs involve alternative use. For example, mining on an NVIDIA GTX 1660 Ti, even with a high Bitcoin price, brings a small income that doesn't even cover the cost of electricity in the USA. If you have a farm of Tesla T4s, you'll owe the electric company about 16 cents a day. Renting farms for cloud gaming or AI training is incomparably more profitable. We are actively integrating vendors who have learned to use idle equipment to reduce your final costs.

We are actively working on optimizing the balance between on-demand and reserved equipment. For example, if your peak CCU is 100 players with an average daily number of 30, and another publisher has a peak of 200 with an average of 50, we calculate the minimum CCU for all our partners launching games with us. Then, we reserve the necessary equipment from vendors a year in advance, thereby obtaining discounts of up to 25-30%. We also reserve a small extra equipment share for peak demand to always ensure stable operation without interruptions.

The Roadmap

By clicking this link, you will find a complete list of more than 70 features and capabilities that we already have and what we can do upon request.

We are currently collaborating with 4 vendors, thanks to which our games are available in 19 regions around the world. The goal is to integrate with another 16 partners by summer to expand our presence. In March, we plan to launch the feature for creating and launching games directly from the back office in a no-code/no-API mode, simplifying the process for everyone, regardless of knowledge level. We want anyone, regardless of their qualifications, to be able to launch a cloud game within 15 minutes.

We have a couple of hypotheses and requests from partners, and we would be happy to hear your opinions and perhaps become a pilot. One of them is to return to the "parts" model. Instead of playing by the hour, we would like to try breaking the game into chapters and offer a cloud solution "by blocks" so as not to compromise the convenience of the gaming experinece for users.

This year, we plan to offer new opportunities to our clients: placing games in the Samsung Gaming Center, Opera GX, and on the X.LA Mall platform with just one click. We are also exploring integration with CrazyGames, Yandex Games, and other portals for HTML5 games, so your mobile and PC games can quickly reach a new audience.

If you are a vendor with cloud technologies and we are not yet in touch – contact us. We are ready to act as your resellers without asking for anything in return. Let us know if you are interested in working with us in the area of Peer-to-Peer cloud gaming when practically any equipment with a video card can be used as a streaming server.

Overall, we have a strong focus on technology and the ability to address a set of use cases for cross-play. For example, play your graphics-heavy game on a mobile device, launch a PC game on a TV, and quickly release a mobile game on the web without porting.

Nick Bondarenko, VP of Product at Xsolla

Join discussion

Comments 0

    You might also like

    We need your consent

    We use cookies on this website to make your browsing experience better. By using the site you agree to our use of cookies.Learn more