 |
SqueezeBrains SDK 1.18
|
|
SqueezeBrains SDK 1.18
|
Information about a computing device. More...
#include <sb.h>
Data Fields | |
| sb_t_device_type | type |
| Device type. | |
| sb_t_frameworks | frameworks |
| Compatible frameworks. More... | |
| char | name [128] |
| Device name. | |
| char | compute_capability [8] |
| Device CUDA Compute capability. More... | |
| int | id |
| Identifier of the device. More... | |
| sb_t_memory_info | memory |
| Memory information about the device and its process. More... | |
| int | available |
| Indicates if the device is available on the PC in use. More... | |
Information about a computing device.
Information about device executing computations of the functions sb_svl_run or sb_project_detection .
| int sb_t_device_info::available |
| char sb_t_device_info::compute_capability[8] |
Device CUDA Compute capability.
CUDA Compute capabilities identifies the features supported by the GPU hardware and used at runtime. An official list of Nvidia GPUs and their associate CUDA Capabilities can be found at the following web page: https://developer.nvidia.com/cuda-gpus.
| sb_t_frameworks sb_t_device_info::frameworks |
| int sb_t_device_info::id |
Identifier of the device.
It is the identifier of the current device among the total devices number of the same type available on the machine. It is always equal to 0 unless more devices of the same type are available (e.g. machine has more GPUs connected). This value must be used to set sb_t_par::devices and sb_t_svl_par::devices used by sb_project_detection and sb_svl_run.
If device type is SB_DEVICE_CPU, it is always 0.
| sb_t_memory_info sb_t_device_info::memory |
Memory information about the device and its process.
1.9.3