A couple of days ago, we shared news that a Life is Strange 2 demo is streamable from a company called Polystream. To better understand Polystream, we emailed them a few questions, which they were kind enough to answer. Check out our full interview below.
Most cloud streaming platforms render graphics on a server, and then stream a video of the application to the client’s computer. Polystream is a cloud streaming platform that renders the application on the client’s computer. Polystream calls this “Command Streaming.”
Would you please describe Command Streaming for our readers?
Command Streaming is an alternative to video streaming that does not require graphics cards in the cloud.
Every game needs to display graphics at some point. With video streaming, the rendering happens on a server in the cloud which is then compressed into video and sent to the client. This process requires a graphics card in the cloud to be fast enough.
With command streaming, however, the rendering happens on the client. When the game issues graphics commands to a graphics API such as DirectX or OpenGL, they are intercepted. These commands are then compressed and sent to the client to render. Notice that the cloud server didn’t need to render anything. This is how Polystream virtualizes the graphics card – and what allows for the server costs to be calculated in cents rather than dollars.
Since Polystream renders graphics client-side, it seems like the burden of graphics processing is on the client. What is the requirement of the graphics card on the client computer? Does their graphics card simply need to meet the minimum specs of the streamed application?
Yes, you’ve got it right, the graphics card simply needs to meet the minimum specs of the streamed application and in today’s world, most devices come standard with an acceptable GPU. But what is also interesting with Polystream is that because we intercept the graphics commands at the driver level, we can optimise them further than even the game can. We do this by batching them and removing redundant commands that a more general game engine would execute.
What are the advantages to Command Streaming over traditional video streaming?
The simple answer is scale. We define true scale as a combination of concurrency, reach and opportunity. If you can’t deliver on all three then you aren’t able to provide scale. Our platform is able to deliver 3D interactive experiences to an unrivalled number of people at the same time. Concurrency is a challenge for traditional solutions and we’ve already outperformed most existing cloud services for our Slush activity last year. Being cloud-neutral, we’re able to operate out of any data centre seamlessly which enables us to help content providers reach audiences worldwide and reduce latency. And finally, as our approach is more cost-effective it enables the growth of new innovative cloud-first experiences and innovative business models supporting the monetization of real-time content. Right now the elephant in the room is that existing solutions are just bad business cases. You might have the content but you won’t have the budget to deliver it. Polystream changes that.
An additional benefit is video streaming solutions compress the output, leading to a loss of visual fidelity; text and images appear blurred. Because we’re delivering commands instead of video we’re also delivering pixel-perfect images rendered by your graphics card so users see applications exactly as they were built.
How much customization is necessary for an application to be compatible with Polystream? Does Polystream require much custom coding from developers?
None. It just works™️
This is one of the biggest benefits of command streaming (apart from the reduced cost, lower latency, bigger scale, and increased quality of course).
Developers do not need to change their code or assets. And they do not need to target a new platform because Polystream runs on the existing Windows build of the game. This is because Polystream not only intercepts graphics commands (command streaming) – but also audio, input, and many other APIs used by the game. In practice, this also means that developers can renew engagement with older games without any effort.
Polystream describes itself as the world’s largest cloud gaming platform. What do you mean by that exactly?
Actually, it’s probably more accurate to describe ourselves as the world’s most scalable cloud platform for 3D interactive content and applications. We believe that real-time 3D content is going to be larger and more impactful than video in the near future from cloud gaming to cloud-first experiences that we haven’t even imagined yet. We’re moving towards a world where the Metaverse is not just an idea but becoming a reality and we are going to be the platform that enables the delivery of all the content needed to power that new wave of experiences at the scale required.
What’s next for Polystream? What features and improvements are coming?
We’re continuing to evolve Polystream so that it can support all types of real-time 3D content and applications. We’ve launched our Polystream Showcase where you can start to see different types of content in action running live from the cloud if you’re on a PC. We recently received an Epic MegaGrant to demonstrate running high-end visual applications from the cloud, and importantly, that it is possible to do this with a wide variety of workloads, from any region, at any scale across the world. Having spent a lot of our early development working on games, primarily a single-window use case, we wanted to demonstrate we could also run high-end visual applications, just as if they were running on your desktop. So we put the Unreal Engine Editor in the cloud which you can try at our showcase here: https://polystream.com/experience-polystream-now/
You’re also able to play Rocket League at our Showcase as well. We’ve just recently added it and as the demo runs for 10mins from the cloud and an average game last 5mins you’ll be able to get two games in without ever downloading the game.