![]() GPU performance recommendations Understanding bandwidth vs. Complex algorithms (i.e inverse kinematics, path-finding). ![]() The following areas are targeted for optimization: Generally, most work in a mixed reality application on the CPU involves doing the "simulation" of the scene and processing your application logic. How to improve your application CPU performance recommendations UnityEngine.XR.XRSettings.renderScale = 0.7f The platform will sample the lower resolution output to build a higher resolution image for rendering on displays. The final image presented on device has a fixed resolution. Unity provides the ability to easily modify the render target resolution of your application at runtime through the XRSettings.renderViewportScale property. Application framerate unchanged, then you're likely CPU Bound.Application framerate increases, then you're likely GPU Bound.The device will render to a smaller texture, then up-sample to display your final image. By reducing the number of pixels to calculate, you'll reduce your GPU load. One way to determine if your app is GPU or CPU bound is to lower the resolution of the render target output. There are many tools that allow you to understand the performance profile and potential bottlenecks in your mixed reality application.īelow are some common tools to help you gather deep profiling information for your application: Use the tools and techniques below to understand where your particular app is bottlenecked. Generally, HoloLens applications will be GPU bound, but not always. It ultimately produces a 2D image to submit to your device's screen. Most commonly handles the graphics pipeline of your application to transform 3D data (models, textures, and so on) into pixels. When your app wants to render an object such as a cube or model, this thread sends a request to the GPU to do the operations. Responsible for submitting your draw calls to the GPU. Responsible for your app logic, including processing input, animations, physics, and other app logic. The three key places where bottlenecks may occur are: There are two primary processors responsible for the work to render your scene: the CPU and the GPU, each handling different aspects of your Mixed Reality app. If your app has an underperforming framerate, the first step is to analyze and understand where your application is computationally intensive. For tips on measuring and improving framerate in the Unity environment, we recommend that you read the performance recommendations for Unity article Understanding performance bottlenecks The framework below outlines best practices for hitting target frame rates. The performant framerate values for each target platform are listed below. Performance must be considered a first class feature for mixed reality development and not a polish task. Holograms will appear unstable and head tracking of the environment will be inaccurate, leading to a poor experience for the user. User experience can be greatly degraded if your application doesn't run at optimal frame rate. This article is an introduction to understanding the significance of performance for your Mixed Reality app.
0 Comments
Leave a Reply. |