We continue to talk about the energy efficiency of mobile devices and Intel applications for Android . Today's post concerns both topics at the same time, because it will be about the Intel Power Monitoring Tool (Intel PMT) program - a tool that allows you to audit the resources consumed both by the system as a whole and by a separate application. Intel PMT may be interesting to curious Android users who want to know everything about their system, but is addressed primarily to developers and serves as a means of diagnosing problems that are not visible to the naked eye.
 The application interface is harsh and not overloaded with any design. The controls are exhausted with just a few buttons that implement the most necessary. You can start the recording process (it goes in the background) and stop it; display the results on the display or the system status bar; send the program log by mail. Finally, a separate button initiates monitoring of a specific application, the name of which, in the best traditions of console utilities, is entered manually.07:40:21 Status Charging BP 406.30 V 3833 I 106 C 692 CPU0 900 27.00% 17.00% 10.00% CPU1 900 42.00% 39.00% 3.00% CpuPStateResidency 0.00% 1.01% 13.13% 85.86% 0.00% CpuCStateResidency 0.01% 3.44% 41.49% 11.09% 43.98% 0.00% 0.00% On 1 Brightness 15.00% WifiTxPackets 0 WifiRxPackets 0 WifiTxBytes 0 WifiRxBytes 0 ModemTxPackets 0 ModemRxPackets 0 ModemTxBytes 0 ModemRxBytes 0 DiskReadBandwidth 0.00 DiskWriteBandwidth 0.00 EmmcReadBandwidth 0.00 EmmcWriteBandwidth 6.00 SdCardReadBandwidth 0.00 SdCardWriteBandwidth 0.00 youtube PID 3858 AppState 83 AppNumOfThreads 48 AppCpuUserUtil 21.00 AppCpuSystemUtil 2.00 AppCpuUtil 23.00

Suppose Mark is an Android mobile application developer. He developed an application for reading books called EReader. He would like to find out if his application has serious problems with power consumption and performance. For this, he decided to use the Power Monitoring Tool. Studying the data, he draws attention to the fact that the total CPU usage and CPU utilization by the application almost reaches 100% without any user actions. He decides to check the profiling data to see if this is a problem. It retrieves the Power Monitoring Tool data and builds CPU utilization graphs by the user application and the system, as shown in the following figure. In this figure, the CPU load by a user application reaches 50%, which means that the application fully uses one core to execute code in user space.
Thanks to data received from the Power Monitoring Tool by the application of the processor by the application, Mark determines that the maximum processor load is at the time of opening the book. Mark guesses that this may be caused by a piece of code opening the e-book. He checks this piece of code and finds an error in it in the form of an infinite loop.
Source: https://habr.com/ru/post/188420/
All Articles