Microsoft claims that disabling Kinect improves Xbox One performance by up to 10%

Quizzical

http://www.dailytech.com/Ditching+Kinect+Boosts+Xbox+Ones+Graphics+Performance+by+10+Percent/article35017.htm

Now, that probably just means that the Kinect was hogging 9% of something, not 9% of everything.  Maybe it was 9% of shaders, or 9% of ESRAM bandwidth, or 9% of something else.  (If you wonder why I'm saying 9% and not 10%, increasing performance by 9% is roughly the inverse of decreasing it by 10%; if you decrease by 10% then increase by 10%, you end up at 99% of original performance.)  So freeing that up to be used by games might typically have little impact, depending on where the bottlenecks were.

But this strikes me as Microsoft admitting that they did something astoundingly stupid a year ago.  Why insist that such resources had to be set aside for dedicated Kinect purposes even when playing games that didn't use the Kinect--meaning, most games?  Giving games extra resources far into the development cycle, or after launch even in many cases, makes it much harder for games to use those extra resources efficiently.

Cleffy

I think Microsoft really wants the Kinect to become a standard input device like the keyboard, mouse, or controller. So by packaging it and dedicating resources to it, it really tells developers they have nothing to lose investing in the Kinect where as with the XBox360, there was no indication that the player would have a Kinect limiting the market. I think that 10% is taken up in all the recognition type things like voice command and face pattern recognition. Then always having it on so the player can send instructions to the OS while they are doing multiple things at once.

I think its a decent move on Microsoft's part to expand gameplay beyond what it has been for the past 3 decades.