Beyond The Cloud: Edge Computing Opens New Frontiers For Online Gaming
“Do we need a next generation of game consoles?”
Guest post by Gaming Industry veteran Justin Keeling who has held senior leadership roles at Fox Interactive, Disney, and many other amazing industry leaders.
GameSpot recently polled their readers and asked them the following question:
“Do we need a next generation of game consoles?”
With 5G on horizon and end-to-end fiber optic pipelines, how close are we to a future where gamers can just stream everything?
75% of readers chose: “Not remotely close”.
The skepticism is understandable. Since 2009, cloud gaming platforms like OnLive and Gaikai have been hyped as the disrupters of the traditional gaming console. I played the first OnLive demo at GDC 2009: Crysis running on a mid-range laptop. Most gaming desktops couldn’t run the game at the time and OnLive felt like black magic.
Outside of the demo environment - and in the real world- it couldn’t live up to the hype. Neither OnLive or Gaikai still operate, although some of their core networking technologies made it into current platforms like PlayStation Now.
But even a decade later nothing yet offers the invisible controller-to-cortex response that most games require.
Edge computing for Gaming
Over the next 18 months, the game streaming experience is set for a resurgence. Microsoft’s xCloud and Google’s Project Yeti have fired the first shots by leveraging their parent companies’ end-to-end ecosystems to build platforms designed around streaming. Amazon and PlayStation are also cooking their own solutions. And all of the top five game publishers are betting heavily on a game streaming future.
What’s different today, is that an industry-wide critical mass of telcos, CDNs, hardware makers, tool providers, content creators – and crucially– a new generation of deep tech researchers are working in coalition to finally make game streaming a reality.
Edge computing is one of the jigsaw pieces needed to make this all work. Simply put; the Edge is a post-cloud network architecture that is hyper-local, decentralized, and much more efficient than the current, monolithic cloud infrastructure that is behind the current digital ecosystem. And the implications for gaming go far beyond just streaming latency.
I’ve been advising Macrometa for a little over a year to understand the implications of their offering to the gaming industry, and I’m excited that Macrometa is attacking one of the most significant parts of the problem at the intersection of gaming and edge computing. This holds the promise of radically transforming the experience of both gamers and game developers. Macrometa has recently emerged from stealth with its planet scaling serverless edge cloud – the Global Edge Fabric. I had an opportunity to speak with Joshua Petty, Macrometa’s Vice President of Products about their potential impact on gaming. He quickly agreed that edge computing offers incredible potential to easily create this new breed of streamed games.
“The team realized early on that there were many benefits to relocating game processing, logic, and analytics closer to the end user. Burying the infrastructure in traditional cloud architectures and databases is costly and severely limits the ability to create truly global, immersive environments for all users. What was missing was a powerful data and messaging platform at the edge that allowed developers to quickly create applications and games using a pre-built and flexible backend for managing dynamic data and content. So we decided to build it as a simple service for developers to consume. Moving everything closer to the end user was the obvious approach to resolve issues of performance and scale.”
“We also realized that improving the experience of developers required more than just providing a global multi-model database or data platform. We went to work at making the developer experience simple and seamless, combining a flexible and globally scaled edge database with serverless functions, containers, Streams that unify pub-sub messaging, and the ability to expose your data models and content as GraphQL or RESTful APIs automatically. The result is a simple backend delivered as a service supporting Unity and Unreal engines, along with java, Python, and other common web languages and runtimes.”
This type of approach could have a significant impact in the near term on some of the bigger challenges faced when attempting to build a ‘streaming everything’ game experience.
Let’s look at some examples:
1. Seamless cross-platform play
Millions of gamers play Fortnite, but the experience is segmented across different platforms and multiple devices. Recently, progress has been made by stitching together some elements of these ecosystems (thanks Microsoft, Nintendo, Epic).
However, Fortnite is a special case. It’s the world’s biggest game, the users have spoken, and “stitched” is the operative word here. PlayStation is still holding out. For most other games, forget it.
Users clearly want the Netflix & Spotify account experience. Cueing a track at home on an Echo, transitioning seamlessly to phone during a commute, and back onto a laptop at work.
The decentralized architecture at the core of Edge Computing, and platform providers like Macrometa that eliminate backend complexity for developers will help to enable this same seamless user experience for high bandwidth games.
2. Next-Generation Match-Making & User Management
There are plenty of back-end-services out there for multiplayer gaming. If you’re looking for the basics of match-making, leaderboards, and friends lists you’re already covered.
Larger game communities are magnitudes more demanding and require custom built tools – entire studios dedicated to live ops management and custom network architecture.
For example, Activision subsidiary Demonware helps build the multiplayer architecture for millions of concurrent players throwing mountains of real-time data at games like Call of Duty and Destiny 2. Under the current limitations of a centralized cloud/CDN internet architecture, it is a miracle these teams have made it work as well as it (usually) does.
But often it doesn’t work well, especially at scale. Destiny 2’s matching has been the source of community frustration for years. The game’s sluggish inventory systems are constantly lagging. Meanwhile, the PlayStation Network achievement ecosystem is bewilderingly slow to sync. Legacy cloud-based database architectures simply can’t keep up with increasing data and state management demands for much longer.
We need a better core solution for this. As we transition into the upcoming generation with more ambitious game designs, more users on the network, and more data flowing through that network, systems built on hyper-local Edge architecture can dramatically improve match-making accuracy and multiplayer user experiences across the board.
3. AR Gaming that works for Everyone
As much as we love Pokémon Go, the sheer volume of users and location data made for a laggy launch experience. It still gets sluggish around high data focal points.
Edge Computing’s architecture is tailor made for a new generation of challenging, high bandwidth AR games. Distributed platforms like Macrometa let developers synchronize player activity in much smaller groups based on location data, and then can automatically trigger functions that respond to local gameplay actions on a peer-to-peer basis.
The states of all in-game entities are also stored in real time at Edge locations much closer to the user than current centralized cloud servers, letting developers execute much more complex AR simulations, and with much less lag.
4. New frontiers in Real-time Data driven gameplay
As Edge distributed databases can sync local game world and player states in real-time, this also has huge implications for game design. If you’re designing experiences with a fixed, or intermittently updated version of what the game world is and what the player’s doing in it, you’re limited to a much less dynamic set of options
Edge Computing also manages data not just in real-time but in context– designers can use this contextual data to create custom experiences based on actual player progression. That opens up exciting new possibilities for truly persistent game play in online worlds, while actually lowering the processing and network overhead on a gamer’s console or device back home.
As we approach gaming’s next generation – our industry’s 10th– we collectively have an opportunity to harness the potential of Edge computing. The applications promise not only to solve age-old problems, but also unlock new experiences for developers increasingly frustrated by an infrastructure built for an older paradigm. 10 years on from the OnLive experiment, it’s time to embrace that future.
Macrometa is a global platform created from the ground-up on the Edge, powering the next generation of planetary scale games and apps. To arrange a demo at GDC or Google Next, contact: firstname.lastname@example.org or email@example.com for interviews and media relations.