LKT - MODULO + APK

LKT

LKT  APK: {BAIXAR AQUI}

LKT MODULO:  {BAIXAR AQUI}

Introduction

LKT aims to achieve a fair balance between power consumption and performance. Instead of tuning the parameters manually, LKT adopts Project WIPE open-source interactive parameters for the mainstream SOCs that are generated by machine learning AI via simulation and heuristic optimization algorithms.

How does it work?

WIPE v2 focuses on meeting performance requirements when interacting with APP, while reducing non-interactive lag weights, pushing the trade-off between fluency and power-saving even further. QTI Boost Framework, which must be disabled before applying optimization, is able to dynamically override parameters based on perf hint. This project utilizes the QTI Boost Framework and extends the ability to override custom parameters. When launching APPs or scrolling the screen, applying more aggressive parameters to improve response at an acceptable power penalty. When there is no interaction, use conservative parameters, use small core clusters as much as possible, and run at a higher energy efficiency OPP under heavy load.