使用者工具

網站工具


jetson_agx_orin_development

差異處

這裏顯示兩個版本的差異處。

連向這個比對檢視

Both sides previous revision 前次修改
下次修改
前次修改
jetson_agx_orin_development [2024/03/19 17:42]
don [Modified]
jetson_agx_orin_development [2024/04/16 15:34] (目前版本)
don [AGX Stable Diffusion]
行 13: 行 13:
 Before JetPack 5.1.1\\ Before JetPack 5.1.1\\
 Jetson Orin module to be emulated Flashing command\\ Jetson Orin module to be emulated Flashing command\\
-Jetson AGX Orin 64GB            sudo ./flash.sh jetson-agx-orin-devkit mmcblk0p1\\ +Jetson AGX Orin 64GB\\            ''sudo ./flash.sh jetson-agx-orin-devkit mmcblk0p1''\\ 
-Jetson AGX Orin 32GB            sudo ./flash.sh jetson-agx-orin-devkit-as-jao-32gb mmcblk0p\\ +Jetson AGX Orin 32GB\\            ''sudo ./flash.sh jetson-agx-orin-devkit-as-jao-32gb mmcblk0p''\\ 
-Jetson Orin NX 16GB            sudo ./flash.sh jetson-agx-orin-devkit-as-nx16gb mmcblk0p1\\ +Jetson Orin NX 16GB\\            ''sudo ./flash.sh jetson-agx-orin-devkit-as-nx16gb mmcblk0p1''\\ 
-Jetson Orin NX 8GB            sudo ./flash.sh jetson-agx-orin-devkit-as-nx8gb mmcblk0p1\\ +Jetson Orin NX 8GB\\            ''sudo ./flash.sh jetson-agx-orin-devkit-as-nx8gb mmcblk0p1''\\ 
-Jetson Orin Nano 8GB*            sudo ./flash.sh jetson-agx-orin-devkit-as-nano8gb mmcblk0p1\\ +Jetson Orin Nano 8GB*\\            ''sudo ./flash.sh jetson-agx-orin-devkit-as-nano8gb mmcblk0p1''\\ 
-Jetson Orin Nano 4GB            sudo ./flash.sh jetson-agx-orin-devkit-as-nano4gb mmcblk0p1\\+Jetson Orin Nano 4GB\\            ''sudo ./flash.sh jetson-agx-orin-devkit-as-nano4gb mmcblk0p1''\\
  
  
行 27: 行 27:
 sudo ./flash.sh jetson-agx-orin-devkit internal sudo ./flash.sh jetson-agx-orin-devkit internal
 </code> </code>
 +
 +==== Build Kernel ====
 +<code>
 +export CROSS_COMPILE_AARCH64_PATH=$HOME/l4t-gcc/
 +export CROSS_COMPILE_AARCH64=$HOME/l4t-gcc/bin/aarch64-buildroot-linux-gnu-
 +cd ~/workspace/NX/Linux_for_Tegra/source/public
 +mkdir kernel_out
 +./nvbuild.sh -o $PWD/kernel_out
 +cd ../../
 +./fl.sh
 +</code> 
  
 ==== Install Vscode on Target Device ==== ==== Install Vscode on Target Device ====
行 122: 行 133:
 </code> </code>
  
-=== 3. https://catalog.ngc.nvidia.com/orgs/nvidia/resources/deepstream ===+=== 3. Get deepstream-6.3_6.3.0-1_arm64.deb ===
  
-Get deepstream-6.3_6.3.0-1_arm64.deb for Jetson not to download for x86\\ +[[https://catalog.ngc.nvidia.com/orgs/nvidia/resources/deepstream/files?version=6.3|Download deepstream-6.3_6.3.0-1_arm64.deb ]] 
-<code>sudo apt-get install ./deepstream-6.3_6.3.0-1_arm64.deb</code>+ 
 +<color #ed1c24>Get deepstream-6.3_6.3.0-1_arm64.deb for Jetson not to download for x86</color>\\ 
 +<code>sudo apt install ./deepstream-6.3_6.3.0-1_arm64.deb</code>
  
 === 4. Copy the generated libraries to the deepstream directory: === === 4. Copy the generated libraries to the deepstream directory: ===
行 187: 行 200:
 <code>./gpu_burn 1000</code> <code>./gpu_burn 1000</code>
  
-=== Fan Control ===+==== Fan Control ====
 <code>/etc/nvfancontrol.conf </code> <code>/etc/nvfancontrol.conf </code>
 FAN_DEFAULT_PROFILE cool \\ FAN_DEFAULT_PROFILE cool \\
行 193: 行 206:
 <code>sudo systemctl restart nvfancontrol</code> \\ <code>sudo systemctl restart nvfancontrol</code> \\
  
-=== Power Mode Setting ===+==== Power Mode Setting ====
 Nvidia Power Model Tool to config ID=0 “MAXN” \\ Nvidia Power Model Tool to config ID=0 “MAXN” \\
  
行 200: 行 213:
 <code>sudo jetson_clocks</code> \\ <code>sudo jetson_clocks</code> \\
  
-=== AGX Stable Diffusion ===+==== AGX Stable Diffusion ===
 + 
 +[[https://huggingface.co/CompVis/stable-diffusion-v-1-4-original|Download sd-v1-4.ckpt]] 
 <code> <code>
-https://github.com/chitoku/stable-diffusion+git clone https://github.com/chitoku/stable-diffusion
 cd stable-diffusion cd stable-diffusion
 git checkout jetson git checkout jetson
 +cp ~/Downloads/sd-v1-4.ckpt /home/aopen/stable-diffusion/models/ldm/stable-diffusion-v1/
 ./docker/run.sh ./docker/run.sh
 python3 scripts/txt2img.py --prompt "Robots marching down a street in Japanese city" --plms python3 scripts/txt2img.py --prompt "Robots marching down a street in Japanese city" --plms
 </code> </code>
  
-=== JETSON AGX Orin performance test ===+==== JETSON AGX Orin performance test ====
 https://github.com/NVIDIA-AI-IOT/jetson_benchmarks https://github.com/NVIDIA-AI-IOT/jetson_benchmarks
- 
- 
  
 <code> <code>
行 306: 行 321:
  
 </code> </code>
 +
 +
 +===== Pinmux Changes =====
 +
 +If your carrier board schematic differs from that of the reference carrier board in Jetson Xavier NX Developer Kit, you must change the pinmux configuration applied by the software.
 +
 +To define your board’s pinmux configuration, download the Jetson Xavier NX pinmux table from the Jetson Download Center. Be sure to get the right version of the table for your SOM.
 +
 +The pinmux table is a spreadsheet that:
 +
 +  * Shows the locations and default pinmux settings
 +
 +  * Acts as a data source for software that defines the pinmux settings in the source code or device tree
 +
 +You must customize the spreadsheet for the configuration of your board, then convert the .dtsi file generated by Excel to a .cfg file. For instructions, see the README file at:
 +
 +Linux_for_Tegra/kernel/pinmux/t19x/
 +You must perform the same conversion for gpio.dtsi and padvoltage.dtsi.
 +
 +
 +
 +[[https://developer.nvidia.com/embedded/downloads#?search=Pin%20and%20Function%20Names%20Guide|pinmux table ]]
 +
 +===== GPIO Changes =====
 +If you designed your own carrier board, to translate from SOM connector pins to actual GPIO numbers you must understand the GPIO mapping formula below. The translated GPIO numbers can be controlled by the driver.
 +
 +To check a GPIO number
 +This example procedure checks the GPIO number of signal name SPI0_CS1.
 +
 +  - Open the Jetson AGX Orin pinmux table. (See Pinmux Changes, above.)
 +  - Search the table for SPI0_CS1.
 +  - Confirm that the Customer Usage field contains the GPIO name GPIO3_PZ.07.
 +  - Search PZ.07 using the following command:
 +
 +cat /sys/kernel/debug/gpio | grep PZ.07
jetson_agx_orin_development.1710841366.txt.gz · 上一次變更: 2024/03/19 17:42 由 don