Unverified50% confidenceFactExact time
DCGM's Profiling-level metrics (identified by the DCGM_FI_PROF_ prefix) require the use of GPU hardware performance counters and may be mutually exclusive with certain third-party profiling tools.
1
Sources
50%
Confidence
Long-term
Relevance
9/3/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedDCGM (Data Center GPU Manager) is a GPU management and monitoring tool provided by NVIDIA, widely used in data center environments.69% similarUnverifiedGPU utilization (DCGM_FI_DEV_GPU_UTIL) measures the proportion of time during a sampling period in which at least one kernel is running on the GPU.69% similarUnverifiedIn llama.cpp, the -ngl (number of GPU layers) parameter controls how many layers are placed on the GPU, used to balance speed and model size.65% similarUnverifiedLlama.cpp's -ngl parameter controls how many Transformer layers are loaded into GPU VRAM for inference; if VRAM is insufficient, layers that don't fit are automatically offloaded to the CPU, forming a hybrid inference mode.64% similarUnverifiedNVIDIA's DCGM can collect more fine-grained metrics than nvidia-smi, including SM activity, Tensor Core utilization, DRAM activity, and PCIe/NVLink communication traffic.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/849971API
curl https://kongchang.com/api/v1/knowledge/claims/849971MCP
get_claim(id=849971)