Jetson tx2 性能模式工具 nvpmodel

来源:互联网 发布:蓝莲花网络 编辑:程序博客网 时间:2024/06/03 15:59

Jetson tx2 性能模式工具nvpmodel

    • Jetson tx2 性能模式工具nvpmodel
      • TX2架构图
      • 性能模式列表
      • 用法
      • 举例
      • 参考文献

Jetson Tegra系统的应用涵盖越来越广,相应用户对性能和功耗的要求也呈现多样化。为此NVIDIA提供一种新的命令行工具,可以方便地让用户配置CPU状态,以最大限度地提高不同场景下的性能和能耗。
Jetson TX2由一个GPU和一个CPU集群组成。 CPU集群由双核丹佛2处理器和四核ARM Cortex-A57组成,通过高性能互连架构连接。GPU 是由两个Pascal 架构的SM组成 , 计算能力 6.2,还有一些用来加速特定应用的ASIC电路.架构图如下

TX2架构图

这里写图片描述

性能模式列表

TX2 拥有6个CPU核心和一个GPU,你可以不必自行运行所有性能/功耗来测试最佳的运行状态,因为NVIDIA的新的命令工具Nvpmodel,提供了5种模式。在Jetson TX2上。 下表列出了CPU内核的模式以及正在使用的CPU和GPU的最大频率。

mode Mode Name Denver 2 Frequency ARM A57 Frequency GPU Frequency 0 Max-N 2 2.0 GHz 4 2.0 GHz 1.30 Ghz 1 Max-Q 0 4 1.2 GHz 0.85 Ghz 2 Max-P Core-All 2 1.4 GHz 4 1.4 GHz 1.12 Ghz 3 Max-P ARM 0 4 2.0 GHz 1.12 Ghz 4 Max-P Denver 2 2.0 GHz 0 1.12 Ghz

TX2默认模式是只开4个CPU, 所以你如果要打开其最大性能的话,那就试试nvpmodel ,切换一下模式吧.

用法

nvidia@tegra-ubuntu:~$ sudo nvpmodel [sudo] password for nvidia: Nvidia Power Model Tool Version 1.0.0Usage:    nvpmodel [-h | --help] [--verbose] [-q | --query] [-p | --parse] [-u | --udata]        [-m | --mode <mode>] [-f | --conf <filename>] [-o | --os <android,l4t>]-h, --help:    Print this help info.--verbose:    Enable verbose log.-p, --parse:    Parse the config file only. Recommended to enable verbose log.-m, --mode <mode>:    <mode> is one of the integer POWER_MODEL ID defined in config file. Switch to the specified power mode.-f, --conf:    explicitly specify the config file.                If it is the only option, then it sets the power mode as default mode configured in the file.                This option can be used for developer usage to specify a config file other than the default one.-o, --os <android,l4t>:    Perform OS specific operations for power model settings. Argument is case insensitive.-q, --query:    Query the current power mode.-w, --wait:    delay exectuion by specified amount of seconds.-u, --udata:    specify the absolute path for user data file when set or query power mode.

举例

  1. nvpmodel -m 2切换模式

    ​切换模式到Max-P Core-All

  2. 比如说查看当前模式nvpmodel -q --verbose,并打印信息

    nvidia@tegra-ubuntu:~$ sudo nvpmodel -q --verboseNVPM VERB: parsing done for /etc/nvpmodel.confNVPM VERB: Current mode: NV Power Mode: MAXN0NVPM VERB: PM_CONFIG: DEFAULT=MAXP_CORE_ARM(3)NVPM VERB:   ACTIVE=MAXN(0)NVPM VERB: POWER_MODEL: ID=0 NAME=MAXNNVPM VERB: /sys/devices/system/cpu/cpu1/online 1NVPM VERB:   value = 1NVPM VERB: /sys/devices/system/cpu/cpu2/online 1NVPM VERB:   value = 1NVPM VERB: /sys/devices/system/cpu/cpu3/online 1NVPM VERB:   value = 1NVPM VERB: /sys/devices/system/cpu/cpu4/online 1NVPM VERB:   value = 1NVPM VERB: /sys/devices/system/cpu/cpu5/online 1NVPM VERB:   value = 1NVPM VERB: /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq 0NVPM VERB:   value = 345600NVPM VERB: /sys/devices/system/cpu/cpu0/cpufreq/scaling_max_freq 2035200NVPM VERB:   value = 2035200NVPM VERB: /sys/devices/system/cpu/cpu1/cpufreq/scaling_min_freq 0NVPM VERB:   value = 345600NVPM VERB: /sys/devices/system/cpu/cpu1/cpufreq/scaling_max_freq 2035200NVPM VERB:   value = 2035200NVPM VERB: /sys/devices/17000000.gp10b/devfreq/17000000.gp10b/min_freq 0NVPM VERB:   value = 140250000NVPM VERB: /sys/devices/17000000.gp10b/devfreq/17000000.gp10b/max_freq 1300500000NVPM VERB:   value = 1300500000NVPM VERB: /sys/kernel/nvpmodel_emc_cap/emc_iso_cap 0NVPM VERB:   value = 0NVPM VERB: 
  3. 更多例子

    nvpmodel -m 2: switch to POWER_MODEL ID=2 of which settings are defined in the default configuration file.nvpmodel -m 2 -o android: switch to POWER_MODEL ID=2 and perform Android specific operations for power mode.nvpmodel -m 2 -f pm.conf: switch to POWER_MODEL ID=2 of which settings are defined in pm.conf.nvpmodel -m 2 -u /data/status: switch to POWER_MODEL ID=2 and store the active mode as user settings in /data/status.nvpmodel -f pm.conf: read the active mode in user data file and set it as the power mode which is configured in pm.conf.            If user data file does not exist or the active mode value is invalid, set defalut mode instead.nvpmodel -q: print the current power mode.nvpmodel -q --verbose: print the current power mode with verbose info.nvpmodel -p -f pm.conf: parse pm.conf and print the result.

参考文献

https://devblogs.nvidia.com/parallelforall/jetson-tx2-delivers-twice-intelligence-edge/

http://www.jetsonhacks.com/2017/03/25/nvpmodel-nvidia-jetson-tx2-development-kit/

原创粉丝点击