Microsoft: Software update unlocks more GPU bandwidth on Xbox One

Devs can now access 10 percent GPU slice previously reserved for Kinect, system functions

Kyle Orland – Jun 4, 2014 6:20 pm | 117

The system update does not turn the system white... only the employee models do that.

The June system update rolling out to Xbox Ones worldwide this week includes surface-level features like external hard drive support, as we mentioned this morning. But Microsoft says the new firmware will also help developers extract more power from the system’s Graphical Processing Unit (GPU), even though the base hardware in the system is obviously staying the same.

Microsoft didn’t trumpet this news in a press release or blog post, but threw it out there in a tweet from Microsoft’s new executive in charge of Xbox, Phil Spencer: “June #XboxOne software dev kit gives devs access to more GPU bandwidth. More performance, new tools and flexibility to make games better.”

As far back as last October, Microsoft was publicly acknowledging how Kinect and system processing took “a conservative 10 percent time-sliced reservation… for the GPGPU processing for Kinect and for the rendering of concurrent system content such as snap mode.” Back then, the company promised it would be opening up that slice of processing time to game developers in the future in a way that didn’t impact the system’s background performance. That appears to be what has come to pass with the system’s latest software update.

“In June we’re releasing a new SDK making it possible for developers to access additional GPU resources previously reserved for Kinect and system functions,” a Microsoft spokesperson told Ars today. “The additional resources allow access to up to 10 percent additional GPU performance. We’re committed to giving developers new tools and flexibility to make their Xbox One games even better by giving them the option to use the GPU reserve in whatever way is best for them and their games.”

Since the Xbox One launched alongside the PlayStation 4 late last year, much has been made of how multiplatform games tend to run at lower resolutions and frame rates on Microsoft’s system. Microsoft wouldn’t commit to this new GPU accessibility narrowing or eliminating that apparent gap. Instead, the company is simply saying it’s up to developers to make use of the new processing power.

“How developers choose to access the extra GPU performance for their games will be up to them,” the spokesperson said. “The extra GPU will add to games’ performance and definition, but it is up to developers to choose how they want to utilize the extra GPU performance.”

The spokesperson went on to tell Ars that the new SDK “will include new options for how developers can use the system reserve as well as more flexibility in our natural user interface reserve (voice and gesture).” Reading between the lines, this seems to imply that developers may need to eschew the use of Kinect-specific voice and gesture-recognition features in their games to make the most use of the newly available slice of GPU power.

Today’s announcement comes just days before Microsoft will begin selling a $400 version of the Xbox One without a bundled Kinect camera, removing a forced addition that many saw as a weight around the system’s neck. That decision had been in the works since April, when Spencer came on board as Microsoft’s executive in charge of Xbox. It followed months of denials that the Kinect would ever be separated from the Xbox One package.

Back in August, months before the Xbox One hit store shelves, Microsoft upped the clock speed on the GPUs in development kits for the system by roughly seven percent. Just before the system came out, we took a deep dive into what we knew about the relative hardware power of the Xbox One and the PlayStation 4, discovering the systems were largely similar with some differences in the details.