Device Structs
- Provided by: nvidia-cuda-dev (Version: 10.1.243-3)
- Source: nvidia-cuda-toolkit
- Report a bug
#define NVML_VALUE_NOT_AVAILABLE (-1)
#define NVML_DEVICE_PCI_BUS_ID_BUFFER_SIZE 32
#define NVML_DEVICE_PCI_BUS_ID_BUFFER_V2_SIZE 16
#define NVML_DEVICE_PCI_BUS_ID_LEGACY_FMT '%04X:%02X:%02X.0'
#define NVML_DEVICE_PCI_BUS_ID_FMT '%08X:%02X:%02X.0'
#define NVML_DEVICE_PCI_BUS_ID_FMT_ARGS(pciInfo)
#define NVML_NVLINK_MAX_LINKS 6
#define NVML_MAX_PHYSICAL_BRIDGE (128)
enum nvmlBridgeChipType_t
enum nvmlNvLinkUtilizationCountUnits_t
enum nvmlNvLinkUtilizationCountPktTypes_t
enum nvmlNvLinkCapability_t
enum nvmlNvLinkErrorCounter_t
enum nvmlGpuTopologyLevel_t
enum nvmlSamplingType_t { NVML_TOTAL_POWER_SAMPLES = 0,
NVML_GPU_UTILIZATION_SAMPLES = 1,
NVML_MEMORY_UTILIZATION_SAMPLES = 2,
NVML_ENC_UTILIZATION_SAMPLES = 3, NVML_DEC_UTILIZATION_SAMPLES
= 4, NVML_PROCESSOR_CLK_SAMPLES = 5, NVML_MEMORY_CLK_SAMPLES =
6 }
enum nvmlPcieUtilCounter_t
enum nvmlValueType_t
enum nvmlPerfPolicyType_t { NVML_PERF_POLICY_POWER = 0,
NVML_PERF_POLICY_THERMAL = 1, NVML_PERF_POLICY_SYNC_BOOST = 2,
NVML_PERF_POLICY_BOARD_LIMIT = 3,
NVML_PERF_POLICY_LOW_UTILIZATION = 4,
NVML_PERF_POLICY_RELIABILITY = 5,
NVML_PERF_POLICY_TOTAL_APP_CLOCKS = 10,
NVML_PERF_POLICY_TOTAL_BASE_CLOCKS = 11 }
Buffer size guaranteed to be large enough for pci bus id
Buffer size guaranteed to be large enough for pci bus id for busIdLegacy
PCI format string for busId
Value:
(pciInfo)->domain, (pciInfo)->bus, (pciInfo)->deviceUtility macro for filling the pci bus id format from a nvmlPciInfo_t
PCI format string for busIdLegacy
Maximum limit on Physical Bridges per Board
Maximum number of NvLink links supported
Special constant that some fields take when they are not available. Used when only part of the struct is not available.
Each structure explicitly states when to check for this value.
Enum to represent type of bridge chip
Represents level relationships within a system between two GPUs The enums are spaced to allow for future relationships
Enum to represent NvLink queryable capabilities
Enum to represent NvLink queryable error counters
Enum to represent the NvLink utilization counter packet types to count ** this is ONLY applicable with the units as packets or bytes ** as specified in nvmlNvLinkUtilizationCountUnits_t ** all packet filter descriptions are target GPU centric ** these can be 'OR'd' together
Enum to represent the NvLink utilization counter packet units
Represents the queryable PCIe utilization counters
Represents type of perf policy for which violation times can be queried
Enumerator:
Represents Type of Sampling Event
Enumerator:
Represents the type for sample value returned
Generated automatically by Doxygen for NVML from the source code.