1 # Copyright (c) 2018-2019 NVIDIA CORPORATION. All rights reserved.
3 @page dwx_verification Verify Your Installation
5 If you have successfully run NVIDIA SDK Manager or followed the installation steps for your platform, you can find the following folders under <b><installation_folder></b>.
7 /<installation_folder>/driveworks-<version>/ main installation folder
8 /<installation_folder>/driveworks/ symbolic link to /usr/local/driveworks-<version>
9 /<installation_folder>/driveworks-<version>/bin precompiled binary samples
10 /<installation_folder>/driveworks-<version>/data data for samples and tools
11 /<installation_folder>/driveworks-<version>/include symbolic link to targets/<platform>/include
12 /<installation_folder>/driveworks-<version>/lib symbolic link to targets/<platform>/lib
13 /<installation_folder>/driveworks-<version>/samples sources for the samples
14 /<installation_folder>/driveworks-<version>/targets/<platform> DriveWorks libraries and header files
15 /<installation_folder>/driveworks-<version>/tools DriveWorks tools
17 @note The specific location of <b><installation_folder></b> depends on the platform on which it is installed. For details, see the installation section in @ref dwx_devguide_getting_started.
19 ### To easily ensure that all components are correctly installed
21 Execute an NVIDIA<sup>®</sup> DriveWorks sample.
23 1. Navigate to the installation folder.
25 cd <installation_folder>/bin/
27 2. Execute the following command.
31 If the installation is successful, the following information about the platform is printed on the console:
33 nvidia@tegra-ubuntu:/usr/local/driveworks/bin$ ./sample_hello_world
34 *************************************************
35 Welcome to DriveWorks SDK
36 [31-7-2018 9:8:24] Initialize DriveWorks SDK v1.2.227
37 [31-7-2018 9:8:24] Release build with GNU 4.9.4 from v1.2.0-rc6-0-g79beb2a against Vibrante PDK v5.0.10.3
38 [31-7-2018 9:8:24] Platform: Detected Drive PX2 - Tegra A
39 [31-7-2018 9:8:24] TimeSource: monotonic epoch time offset is 1532010806746835
40 [31-7-2018 9:8:24] TimeSource: PTP ioctl returned error. Synchronized time will not be available.
41 [31-7-2018 9:8:24] TimeSource: Could not detect valid PTP time source at 'eth0'. Fallback to CLOCK_MONOTONIC.
42 [31-7-2018 9:8:25] Platform: number of GPU devices detected 2
43 [31-7-2018 9:8:25] Platform: currently selected GPU device discrete ID 0
44 [31-7-2018 9:8:25] SDK: Resources mounted from /usr/local/driveworks-1.2/data/resources
45 [31-7-2018 9:8:25] SDK: Create NvMediaDevice
46 [31-7-2018 9:8:25] SDK: Create NvMediaIPPManager
47 [31-7-2018 9:8:25] egl::Display: found 2 EGL devices
48 [31-7-2018 9:8:25] egl::Display: use drm device: drm-nvdc
49 Context of Driveworks SDK successfully initialized.
51 GPU devices detected: 2
52 [31-7-2018 9:8:25] Platform: currently selected GPU device discrete ID 0
53 [31-7-2018 9:8:25] Platform: currently selected GPU device discrete ID 0
54 ----------------------------------------------
55 Device: 0, DRIVE PX 2 AutoChauffeur
56 CUDA Driver Version / Runtime Version : 9.2 / 9.2
57 CUDA Capability Major/Minor version number: 6.1
58 Total amount of global memory in MBytes:3839.94
59 Memory Clock rate Ghz: 3003000
60 Memory Bus Width bits: 128
61 L2 Cache Size: 1048576
62 Maximum 1D Texture Dimension Size (x): 131072
63 Maximum 2D Texture Dimension Size (x,y): 131072, 65536
64 Maximum 3D Texture Dimension Size (x,y,z): 16384, 16384, 16384
65 Maximum Layered 1D Texture Size, (x): 32768 num: 2048
66 Maximum Layered 2D Texture Size, (x,y): 32768, 32768 num: 2048
67 Total amount of constant memory bytes: 65536
68 Total amount of shared memory per block bytes: 49152
69 Total number of registers available per block: 65536
71 Maximum number of threads per multiprocessor: 2048
72 Maximum number of threads per block: 1024
73 Max dimension size of a thread block (x,y,z): 1024,1024,64
74 Max dimension size of a grid size (x,y,z): 2147483647,65535,65535
75 Maximum memory pitch bytes: 2147483647
76 Texture alignment bytes: 512
77 Concurrent copy and kernel execution: Yes, copy engines num: 2
78 Run time limit on kernels: No
79 Integrated GPU sharing Host Memory: No
80 Support host page-locked memory mapping: Yes
81 Alignment requirement for Surfaces: Yes
82 Device has ECC support: Disabled
83 Device supports Unified Addressing (UVA): Yes
84 Device PCI Domain ID: 0, Device PCI Bus ID: 4, Device PCI location ID: 0
85 Compute Mode: Default (multiple host threads can use ::cudaSetDevice() with device simultaneously)
89 [31-7-2018 9:8:25] Platform: currently selected GPU device integrated ID 1
90 [31-7-2018 9:8:25] Platform: currently selected GPU device integrated ID 1
91 ----------------------------------------------
92 Device: 1, NVIDIA Tegra X2
93 CUDA Driver Version / Runtime Version : 9.2 / 9.2
94 CUDA Capability Major/Minor version number: 6.2
95 Total amount of global memory in MBytes:6401.58
96 Memory Clock rate Ghz: 1600000
97 Memory Bus Width bits: 128
99 Maximum 1D Texture Dimension Size (x): 131072
100 Maximum 2D Texture Dimension Size (x,y): 131072, 65536
101 Maximum 3D Texture Dimension Size (x,y,z): 16384, 16384, 16384
102 Maximum Layered 1D Texture Size, (x): 32768 num: 2048
103 Maximum Layered 2D Texture Size, (x,y): 32768, 32768 num: 2048
104 Total amount of constant memory bytes: 65536
105 Total amount of shared memory per block bytes: 49152
106 Total number of registers available per block: 32768
108 Maximum number of threads per multiprocessor: 2048
109 Maximum number of threads per block: 1024
110 Max dimension size of a thread block (x,y,z): 1024,1024,64
111 Max dimension size of a grid size (x,y,z): 2147483647,65535,65535
112 Maximum memory pitch bytes: 2147483647
113 Texture alignment bytes: 512
114 Concurrent copy and kernel execution: Yes, copy engines num: 1
115 Run time limit on kernels: No
116 Integrated GPU sharing Host Memory: Yes
117 Support host page-locked memory mapping: Yes
118 Alignment requirement for Surfaces: Yes
119 Device has ECC support: Disabled
120 Device supports Unified Addressing (UVA): Yes
121 Device PCI Domain ID: 0, Device PCI Bus ID: 0, Device PCI location ID: 0
122 Compute Mode: Default (multiple host threads can use ::cudaSetDevice() with device simultaneously)
123 Concurrent kernels: 1
126 [31-7-2018 9:8:25] SDK: Release NvMediaDevice
127 [31-7-2018 9:8:25] Driveworks SDK released
128 [31-7-2018 9:8:25] SDK: Release NvMedia2D
129 Happy autonomous driving!
133 @note For a full list of samples that can be run out of the box, see @ref dwx_samples_section.
134 @note For detailed instructions on building the samples from source see @ref dwx_samples_building.